Authors: Ge Wang
Abstract: Here we introduce ELECTRIC (Evidential Learning-Enhanced CT Reconstruction via Iterative Correction), a physics-guided Bayesian formulation. An evidential neural network provides an image proposal and an error-predictive epistemic-uncertainty surrogate. The latter is converted into an adaptive precision field and inserted into a Poisson-weighted MAP update. The resulting image-evidence-precision-reconstruction loop treats prior confidence as a learned state variable of iterative reconstruction. In addition to the formulation and theoretical analysis, we report two simulation studies on image slices from the AAPM Mayo Clinic Low-Dose CT dataset: a mechanism-validation pilot using transparent surrogate estimators, and a feasibility study in which a trained Normal-Inverse-Gamma evidential network drives the full closed loop. On held-out patients, the learned prior mean reduces reconstruction error by roughly 70 percent relative to filtered back-projection, the learned epistemic uncertainty is error-predictive and supports selective trust, and the physics-guided update restores measurement consistency while the adaptive-precision reconstruction matches or exceeds a validation-tuned fixed prior and remains markedly more robust to prior-strength misspecification. Together these results demonstrate the complete ELECTRIC closed-loop pipeline, while identifying formal uncertainty calibration and joint training as the principal directions for future work.
Authors: Adrian Urba\'nski, Gabriel della Maggiora, Artur Yakimovich
Abstract: Generative models learn the statistical properties of their training data, so high-quality generation depends on clean and representative datasets. In scientific imaging, acquisition often yields noisy measurements, while collecting clean references can be costly, impractical or even unattainable. Training directly on these measurements results in a model that reproduces the corrupted data. This can be circumvented by learning the clean population distribution directly from the noisy data. Conditional flow matching (CFM) combines a simple regression objective with stable training, efficient sampling, and strong image-generation performance, making it a natural framework for this setting. We introduce Noise-Robust Conditional Flow Matching (NR-CFM), an unconditional generator that learns from one corrupted observation per image. NR-CFM provides a closed-form clean endpoint correction for additive white Gaussian noise and learns a data-driven correction for general Gaussian corruptions with more complex covariance structure. Across the evaluated corruption settings, NR-CFM outperforms NR-GAN in most cases and remains competitive with Ambient Diffusion in the high-noise regime. We further evaluate NR-CFM on scientific data at signal-to-noise ratios as low as $0.001$, where it generates plausible particle images from severely corrupted measurements.
Authors: Yehui Yang, Bo Zhao, Junzhe Cao, Hui Ma, Yue Sun, Wenjin Wang, Zitong Yu
Abstract: Remote photoplethysmography (rPPG) enables non-contact heart-rate estimation from facial videos, but its weak physiological signal is easily corrupted by motion, illumination changes, occlusion, skin-appearance variation, and device noise. Existing rPPG methods typically rely on a single model to directly predict heart rate or recover pulse waveforms, while different strong estimators may produce conflicting yet individually plausible candidates for the same video. To resolve these conflicts, we propose PhysAgent, an inference-time multi-agent candidate-verification framework. Unlike direct prediction approaches, PhysAgent neither trains a new base rPPG model nor asks Multimodal Large Language Models (MLLMs) to output heart rate directly. In contrast, it treats outputs from multiple base estimators as physiological hypotheses to be verified and uses a lightweight 4B MLLM, Qwen3-VL-4B, to drive multi-agent reasoning over video conditions, signal reliability, and candidate disagreement. A deterministic physiological verifier checks the fusion proposal, and a reproducible numerical fusion process produces the final heart rate. Experimental results on multiple public rPPG benchmarks show that PhysAgent improves fusion stability and reliability across different datasets and source-domain settings, while avoiding the irreproducibility and physiological inconsistency of direct MLLM prediction or unconstrained ensemble fusion. The code will be released soon.
Authors: Yi Cui, Zilin Wang, Yijie Xu, Qianyi Cai, Huizai Yao, Shuai Jiang, Bingzhuo Zhong, Hui Xiong
Abstract: Construction-safety models must handle concrete deployment risks, such as a worker standing near a scaffold edge without guardrails, rather than only recognize common objects in curated images. Yet real inspection archives are redundant, long-tailed, and collected across changing sites and months. We introduce SafeBuild-Bench, a metadata-driven benchmark for evaluating multimodal large language models on construction safety under realistic temporal and site variation. It is mined from 100K+ industrial image-text records and contains 3,314 task instances from over 3,000 expert-verified images, covering multiple-choice hazard identification and free-form hazard description. To make expert verification scalable, we develop GEMS, a graph-enhanced multimodal selection pipeline that combines a proxy-model confusion signal with graph-based diversity to identify informative candidates from redundant streams. On public instruction-tuning data, GEMS-selected subsets preserve robustness-oriented performance under small data budgets. On SafeBuild-Bench, current MLLMs remain far from reliable construction-safety understanding, with the best overall score near 60. We release the benchmark, evaluation scripts, and GEMS codebase at https://github.com/safebuild/gems.
Authors: Bilel Guetarni, Feryal Windal, David Pasquier, Halim Benhabiles
Abstract: The rapid emergence of 3D CT foundation models has opened new avenues for predictive modeling from CT imaging, offering a compelling alternative to traditional radiomics which is known to suffer from reproducibility issues and sensitivity to acquisition protocol variations. Yet, as these models grow in availability, a critical need arises to evaluate how well their learned representations generalize across diverse clinical settings and whether adaptation to specific downstream tasks is necessary to unlock their full potential. To address these questions, we benchmarked several 3D CT foundation models for predicting recurrence-free survival in head and neck cancer across two public datasets totaling 3,644 patients, evaluating various adaptation strategies and modality fusion mechanisms. Our findings reveal persistent difficulty in identifying features that generalize consistently across different imaging distributions, as evidenced by significant performance drops on external validation cohorts. Ultimately, the integration of imaging features with clinical data remains the most accurate approach for prognostic prediction, though achieving universal generalization across varied clinical contexts continues to represent a substantial challenge for the current generation of models.
Authors: Federica Torrisi, Claudia Corradino, Alessandro Grilli, Tommaso Catuogno, Mattia Verducci, Elisabetta Paladino, Luigi Giannelli, Alessandro Sebastianelli
Abstract: Recent advances in quantum computing are opening new possibilities for Earth Observation (EO) data analysis. Quantum machine learning (QML) approaches offer novel ways to process information by exploiting quantum phenomena such as superposition and entanglement. These capabilities have motivated the exploration of whether quantum-enhanced models can address long-standing challenges in satellite remote sensing, where complex spectral and spatial signals often require sophisticated feature extraction. Among various fields of application, EO data allow the global monitoring of volcanic clouds and are crucial for aviation safety, hazard assessment, real-time eruption response, and evaluation of volcanic impacts on climate. Yet accurate detection of volcanic clouds remains difficult due to their similarity with meteorological clouds, the variability of eruption signatures, and the coarse spectral sampling of geostationary sensors. In this work, the potential of hybrid quantum convolutional neural networks (QCNNs) for the classification of satellite images containing volcanic clouds was investigated. These architectures integrate quantum computational layers into a classical convolutional framework. Two QCNN variants (with 2 and 4 qubits) have been considered to evaluate their ability to classify a dataset of SEVIRI images, including scenes with volcanic clouds (composed of ash, $SO_2$, or mixed components) as well as non-volcanic backgrounds. Finally, the performance of the hybrid QCNN models was compared with that of purely classical architectures.
Authors: Qinghui Liu, Jon Andr\'e Ottesen, Atle Bj{\o}rnerud, Kyrre Eeg Emblem
Abstract: Rigorous dataset partitioning is a foundational, yet frequently overlooked, prerequisite for reliable deep learning in longitudinal medical imaging. Naively shuffling small clinical cohorts routinely introduces covariate shifts and temporal sampling imbalances across training, validation, and test subsets, exposing downstream models to out-of-distribution evaluation. We address this vulnerability with an auditable Tripartite Dataset Analytics Framework that systematically characterizes spatial grid integrity, multi-parametric intensity fingerprints, and longitudinal temporal trajectories, quantifying the heavy-tailed feature dispersion and irregular, episodic sampling intervals typical of real-world clinical cohorts. Building on this characterization, we formalize an unsupervised spatio-temporal cohort-balancing standard operating procedure (SOP) that combines elbow-optimized K-means clustering over a standardized, six-dimensional joint intensity-temporal feature space with intra-cluster proportionate stratified sampling. On a longitudinal, contrast-enhanced $T1$-weighted brain MRI cohort (N=149), the protocol reduces the maximum cross-subset intensity bias from 34.1% under conventional random shuffling to under 2.1%, while aligning longitudinal follow-up intervals closely around the population mean. Monte Carlo stress testing across ten random seeds and three split configurations confirms that this alignment remains tightly bounded, in clear contrast to the substantial variability of random partitioning. The resulting protocol offers a reproducible, generalizable procedure for cohort engineering in variable-length longitudinal clinical imaging workflows.
Authors: Nevio Dubbini, Daniel P. van Helden, Claudia Sciuto, Martina Naso, Arthur Leck, Clement Joubert, Heeli C. Schechter, Remy Chapoulie, Gabriele Gattiglia
Abstract: This paper presents a multimodal machine-learning framework for calibration monitoring, quality assessment, and adaptive acquisition support in archaeological digitisation workflows. The proposed approach operates across photogrammetric 3D reconstruction, hyperspectral imaging, X-ray fluorescence spectroscopy, and Raman spectroscopy through a unified pipeline combining deterministic quality indicators, statistical feature representations, machine-learning classification, anomaly detection, and explainable artificial intelligence (XAI). Rather than replacing instrument-level calibration, the framework introduces an additional algorithmic layer that evaluates whether acquisitions are statistically consistent, physically plausible, and suitable for downstream multimodal integration. For each sensing modality, acquisitions are represented through structured feature spaces encoding geometric, spectral, spatial, and statistical properties. These representations are used to identify degradation patterns such as reconstruction artefacts, illumination inconsistencies, spectral distortions, detector instability, baseline fluctuations, and low signal-to-noise conditions. Supervised and unsupervised learning methods are combined with XAI techniques to support both automatic discrimination between acceptable and problematic acquisitions and interpretation of the underlying causes of degradation. The framework additionally supports adaptive feedback and resource-aware acquisition strategies by linking feature-space deviations to acquisition-level corrective actions. Experimental results obtained on multimodal archaeological datasets demonstrate that the proposed methodology captures meaningful acquisition variability and enables robust quality assessment across heterogeneous sensing modalities.
Authors: Yu Guan, Mingyu Hu, Jiale Hu, Zhuoxu Cui, Dong Liang, Qiegen Liu
Abstract: Accelerated magnetic resonance imaging (MRI) aims to recover the k-space signal from acquired measurements, where accurate estimation of missing samples is essential for high-fidelity reconstruction. Existing k-space reconstruction methods estimate missing samples through interpolation operators or structure priors defined on discrete sampling grids. Although these formulations effectively exploit local interpolation relationships and global k-space redundancy, they reconstruct only discrete frequency coefficients and therefore do not explicitly model the underlying continuous signal. To overcome this limitation, we propose K-space Gaussian Representation (KGR), the first explicit continuous representation formulated directly in the native k-space domain. Rather than estimating unknown samples on discrete grids, KGR parameterizes the continuous signal using Gabor-Gaussian primitives with shared spatial geometry, yielding a compact representation that naturally preserves inter-coil correlations. Because unconstrained continuous fitting does not necessarily satisfy the intrinsic structural properties of multi-coil signal, the estimated representation is projected onto a low-rank manifold to enforce the algebraic constraints arising from smoothly varying phase and coil redundancy. A frequency-adaptive fitting strategy accommodates the heterogeneous characteristics of different k-space regions. Comprehensive validation across multiple datasets and sampling schemes shows consistent improvements over representative reconstruction baselines in both quantitative metrics and visual quality. These results suggest that explicit continuous parameterization of native k-space provides a principled framework for integrating continuous signal modeling with structured low-rank reconstruction.
Authors: Vahidin Hasic, Chao Wang, Luis C. Garcia-Peraza-Herrera, David Watson, Senka Krivic
Abstract: Multimodal large language models (MLLMs) increasingly support high-stakes decision making by combining complementary information from images and text. While existing explainability methods identify influential image regions or text tokens, they cannot answer a fundamental question: which modality drives a prediction? Consequently, a model may produce the correct output while relying on the wrong source of evidence, masking shortcut learning and unsafe reasoning. We formulate modality attribution as a complementary explainability objective for multimodal foundation models and propose Counterfactual Modality Attribution (CMA), the first framework for quantifying modality-level contributions in MLLMs. CMA generates image-only, text-only, and joint multimodal counterfactuals using coupled diffusion priors and converts them into principled modality attribution scores through a cooperative game-theoretic formulation based on Shapley values. We evaluate CMA on controlled synthetic benchmarks with known ground-truth modality reliance and on a real-world multimodal clinical dataset. CMA correctly identifies the decision-driving modality in 98% of controlled cases and consistently outperforms baselines, revealing failures of cross-modal reasoning that remain invisible to predictive accuracy alone. Our results establish modality attribution as a complementary dimension of explainability beyond feature attribution, providing a principled framework for auditing multimodal foundation models in safety-critical applications.
Authors: Feixiang Liu, Qiang Qiu, Hao Zhang, Xinyue Wang
Abstract: Visual-token pruning is usually judged by answer quality at a fixed retention budget. For text-rich multimodal large language models (MLLMs), this protocol can miss a distinct failure: an answer remains correct even when no retained token is locally traceable to the small OCR region that supports it. We turn this blind spot into an evidence-risk audit that couples answer behavior with geometric token-origin provenance, interventions, and realized cost; transparent training-free selectors isolate controlled operating points. On locked image-disjoint confirmation, Qwen Target at 30% retention has observed accuracy 0.786 versus 0.783 for Full (paired image-cluster difference +0.003, 95% CI [-0.014, +0.020]), yet same-budget Target, Random, and Grid retain sharply different positive-support coverage: 0.620, 0.270, and 0.318. Across Qwen3-VL-8B, LLaVA-1.5-7B, and InternVL3.5-8B, matched controls, interventions, detector tests, and external methods reveal model-specific quality-risk-traceability frontiers that accuracy alone does not expose. Materialized prefixes yield up to 4.32x batch-prefill speedup and 76.4% lower incremental peak memory; full-validation TextVQA and DocVQA further show that favorable target-verification points do not imply task-general compression. Visual-token pruning should therefore report surviving spatial provenance and realized cost alongside quality and compression.
Authors: Saif U Din, Muhammad Ahsan Hussain, Radu Timofte, Dmitry Ignatov
Abstract: Deploying convolutional neural networks generated by large language models (LLMs) on real mobile hardware requires more than GPU validation accuracy: INT8 TensorFlow Lite export, delegate selection, and on-device latency jointly determine whether a model is usable. We present an automated mobile deployment pipeline that closes the loop from QLoRA fine-tuning of an architecture-generating LLM through GPU evaluation, INT8 export, and physical-device benchmarking to gated augmentation of the training corpus. The pipeline is fully scripted and runs cycle-by-cycle without manual intervention, with resume support after interruptions. We evaluate the same frozen protocol on two benchmarks, CIFAR-10 and CIFAR-100, on a Samsung SM-P613 tablet (seed 42, 20 models per cycle, cycles 0-6). On CIFAR-10, cycle 1 is gate-accepted and improves the mobile deployment score approximately 25.6x over the baseline with a mean quantized accuracy of 46.9%; later cycles raise GPU accuracy but fail the non-decreasing mobile gate. On CIFAR-100, the pre-QLoRA baseline retains the best mobile score; iterative rounds improve GPU accuracy (up to 26.2%) yet cannot surpass cycle 0 on-device, and the training pool stalls at 19 examples after the first accepted round. Together, the two studies show that closed-loop GPU fine-tuning does not guarantee monotonic mobile gains, especially on harder classification tasks, and that multi-dataset, on-device measurement is needed to stress-test deployment objectives. We release per-cycle metrics with 95% confidence intervals, all figures, and complete reproduction commands.
Authors: Rongxiang Zhang, Songhua Liu
Abstract: Long-form and real-time talking-head generation remains challenging due to a latency-quality trade-off: inefficient multi-step diffusion prohibits streaming generation, whereas real-time autoregressive approaches suffer from error accumulation and identity drift. To address this drawback, we propose LeapTalk, a novel framework that achieves stable and real-time talking-head generation with a single forward step, scaling to arbitrarily long videos. At the heart of our approach lies a single-step bridge distillation scheme. On the one hand, departing from the conventional noise-to-data paradigm, we introduce a data-to-data transport formulation based on a Brownian bridge. Anchored by a persistent reference, this strategy effectively mitigates identity drift and enhances long-term temporal stability. On the other hand, to enable smooth knowledge transfer from a pre-trained diffusion teacher to the student bridge model, we explore a heterogeneous distillation framework with an SNR-aligned time transformation $\Phi(\tau)$, which bridges the functional discrepancy between the two models. Moreover, we propose an audio-driven classifier-free guidance mechanism to maintain fine-grained lip synchronization under extreme step reduction. Extensive experiments demonstrate that our method achieves high-fidelity and temporally consistent video generation with only 1 step at up to 200 FPS, significantly outperforming existing approaches in both efficiency and stability. Project Page: https://zhangrongxiang.github.io/leaptalk-page/
Authors: Arisha Prasain
Abstract: Commercial mapping partnerships are often unavailable in low-resource regions, leaving satellite basemaps stale and motivating synthesis of satellite imagery from independently maintained cartographic data. Existing ControlNet-based diffusion methods typically condition on structural signals like edges or segmentation extracted from the target image itself, assuming the imagery already exists and limiting their use exactly where synthesis matters most. Map-conditioned alternatives add cues like edge detection but omit frequency-domain structure. We propose a ControlNet-based diffusion framework conditioned only on cartographic sources obtainable independently of the target imagery: OpenStreetMap (OSM) raster maps and their stationary wavelet transform (SWT) subbands, a conditioning signal previously unexplored for map-to-satellite diffusion. Two ControlNet adapters, trained separately on the map and wavelet representations atop a frozen Stable Diffusion backbone, are fused via MultiControlNet, jointly drawing on spatial structure and frequency detail without retraining a multi-input model. We evaluate on a new paired map-satellite dataset curated for Nepal, a data-scarce, topographically diverse region, alongside the Pix2Pix maps-satellite benchmark. Combined conditioning wins six of eight metric-dataset comparisons -- SSIM and PSNR on both datasets, plus LPIPS (both Alex and VGG backbones) on ours and ties map-only on both Pix2Pix LPIPS backbones while still edging past wavelet-only there. Wavelet-only takes the lowest FID on both datasets, matching the tradeoff between per-image fidelity and distributional realism. We treat this gap cautiously given our modest test-set sizes and FID's known small-sample bias.
Authors: Aadarsh Agarwal, Kenaish Al Qubaisi, Dirk Englund
Abstract: We present PixCell, a neurosymbolic system in which multimodal agents convert a visually presented photonic component into a parametric program over a small domain-specific language (DSL) of geometric primitives. A system enabling deterministic visual verification renders evaluation asymmetrically cheaper than the generation attempt. While models using multi-seed sampling and iterative revision reach a mean best-turn IoU of only 0.416, multimodal agents through PixCell's interface and verifier consistently exceed 0.9 mean IoU, with scores reaching 0.974 and 0.955 across eight component targets while also satisfying source contracts. These results demonstrate that frontier multimodal agents can reliably understand and render executable parametric representations from visual targets. Using these live parameters, cross-stack studies on an interferometer reconstruct primitive programs that satisfy an 8.0 nm free spectral range target and the original footprint constraint on modeled 220-nm SOI, 400-nm SiN, and 400-nm TFLN stacks. PixCell further carries a paper-derived splitter from visual reconstruction through SOI full-wave simulation, producing symmetric propagation and balanced outputs. Finally, the same executable verifier supplies a training reward and dataset used to train a Qwen3.6-35B-A3B model with LoRA and GRPO without supervised demonstrations. On eight training-excluded paper figures, its mean champion IoU rises from 0.422 after eight initial attempts to 0.491 after three verifier-guided revision rounds. These results therefore establish a controlled framework for measuring, retargeting, and improving visual-to-parametric photonic component design.
Authors: Hoang Thanh Thanh Truong, Charles R. Clark
Abstract: Brain tumor diagnosis is a time-sensitive process in which patients may wait weeks for a finalized pathology report. This problem motivates automated systems that classify tumor subtype from multimodal inputs. This paper details the DS@GT ARC team's work for ImageCLEFmed MEDIQA-CORE 2026 Task~1, Brain Tumor Subtype Classification. The task evaluates three glioma classification problems: Level-1 Molecular Type, LGG vs HGG, and WHO Grade. We combine pre-extracted MRI (NeuroVFM) and histopathology (Prov-GigaPath) embeddings with free-text radiology reports. Our team explored two trimodal fusion architectures, two report encoders (RadBERT and Llama-3.1-8B-Instruct), and a biologically motivated post-processing stage. We achieve a mean macro-F1 of 0.801 under the Fully Multimodal condition, exceeding the organizers' baseline of 0.796 and ranking second among the teams whose code passed verification. Additional evaluation across modality-dropping conditions shows that this advantage depends heavily on the availability of the histopathology modality, and that our system falls behind the baseline when modalities are missing. Our code is available on GitHub at https://github.com/dsgt-arc/imageclef-mediqacore-2026.
URLs: https://github.com/dsgt-arc/imageclef-mediqacore-2026.
Authors: Lisa Ko{\ss}mann, Ralf Bartho, Christoph Redies, Johan Wagemans
Abstract: With rapid growth in the fields of empirical and computational aesthetics we have seen a vast increase in large image datasets annotated for aesthetics. As the image databases differ widely in many respects (e.g., different standards for annotation), it can be tedious to find the dataset that fits one's research needs best. The absence of a centralized open-science search system causes additional problems. Currently, researchers typically share dataset links in papers or on diverse platforms like OSF, GitHub or Dropbox. Manually searching for details like image quality and content often requires downloading all datasets. Therefore, we present the Database Of Datasets for Aesthetics (DODA), an intuitive Web application in which researchers can browse all important datasets for aesthetics research. DODA provides general information about these datasets (size, resolution, type of annotation, number of annotators, etc.) and for many of them also precomputed quantitative image properties. We discuss relevant criteria for selecting a suitable dataset with DODA and illustrate the benefits of reusing datasets. Our approach facilitates collaboration across the fields of empirical and computational aesthetics. Keywords: empirical aesthetics, computational aesthetics, machine learning, image annotation, quantitative image properties, Open Science
Authors: Junhao Chen, Mingjin Chen, Henghaofan Zhang, Minglin Chen, Liaoyuan Fan, Boran Zhang, Saining Zhang, Mingze Sun, Hao Zhao, Ruqi Huang, Zhihao Li, Yufei Wang
Abstract: Pretrained video diffusion models can act as renderers when the desired scene state is already specified by an animated mesh, a camera trajectory, and a reference image. This 4D generative rendering setting raises a representation question: what image-format condition lets a video backbone obey both camera motion and scene-internal animation? We propose DAR, a reference-guided renderer that extends Wan2.2 camera control from Pl\"ucker rays alone to a joint camera-plus-geometry interface. DAR projects a neural 4D G-buffer (tracking, world position, and normal) from the animated mesh and injects it through a widened control adapter while preserving the pretrained image-to-video prior. The central design choice is the pair of tracking and world position. Tracking identifies the persistent surface element that should carry appearance; world position gives its current scene-coordinate state; normal supplies local shape. Depth plus calibrated rays can recover 3D in principle, but depth is a camera-dependent chart in which camera and object motion are mixed. On the 68-case DAR-4D benchmark, LoRA DAR reaches PSNR 23.22, SSIM 0.895, and LPIPS 0.134, improving over off-the-shelf Wan2.2-Depth by 1.54 dB PSNR; a full fine-tune reaches PSNR 25.36 and SSIM 0.917. Matched ablations show that replacing world position by depth reduces PSNR by 1.26--1.55 dB at every checkpoint, supporting tracking+world-position correspondence as a practical 4D rendering condition.
Authors: Daqian Shi, Wei Cao, Xiaoyu Zheng, Lida Shi, Xiaolei Diao, Cedric M John
Abstract: Current deep learning-based character vision studies, e.g., text recognition, character image denoising, and historical text completion, are offering new solutions for learning, managing, and utilizing character resources. However, the performance of these studies peaks only with large and balanced datasets, which is a rarity with real-world character datasets, especially for logographic character languages, e.g., Chinese. The imbalance in data distribution of logographic characters is a common issue due to differences in character usage frequency and new characters being continuously created. In this paper, we propose a novel method for logographic character recognition, which introduces a multi-modal learning approach using visual semantics and contextual semantics of characters. A novel pre-training strategy is designed to enhance deep visual representations, especially for datasets suffering from issues of imbalanced and rare instances, by extracting the contextual semantics of each character from the corresponding language models. We conduct experiments across various datasets to evaluate our character recognition method and further validate the contrastive pre-training strategy by several downstream tasks. Experimental results demonstrate the superiority of our method compared to state-of-the-art methods.
Authors: Satvik Tripathi, Mustafa Ege Seker, Kristian Quevada, Ebubechukwu D Enwerem, Pratham Khandelwal, Emine Meltem, Bera Koca, Shahriar Faghani, Jacinta Arnold, Dania Daye, Tessa S. Cook
Abstract: Vision-language models (VLMs) are increasingly being evaluated for medical imaging, but many available benchmarks emphasize disease classification, report generation, or broad visual question answering rather than the spatial and anatomical reasoning required for radiology. We developed the Spatial Perception and Anatomical Reasoning in Clinical Radiology (SPARC-Rad) Benchmark, a manually curated multimodal benchmark dataset and evaluation pipeline for assessing these capabilities in radiology VLMs. SPARC-Rad includes 300 image-question pairs derived from healthy control imaging studies in The Cancer Imaging Archive (TCIA), spanning CT, MRI, and radiography across the abdomen, chest, breast, neuro, and musculoskeletal categories. Radiology trainees manually designed and annotated questions to evaluate anatomical identification, localization, laterality, regional recognition, device identification, and inter-structure spatial relationships. The evaluation pipeline supports standardized prompting, structured output collection, response normalization, LLM-as-judge grading, human quality review, binary correctness scoring, and subgroup analysis by modality, anatomy, and reasoning type. SPARC-Rad provides a reusable framework for evaluating whether VLMs can provide reasoning for radiologic anatomy as a spatial system, supporting future model development, failure-mode analysis, and pre-deployment assessment.
Authors: Chimdi Walter Ndubuisi
Abstract: Pathology foundation models are reported to encode molecular programmes in tissue morphology, but the evidence is usually a cohort-wide ranked gene list rather than a prediction for a held-out patient. We rebuild such an analysis with the patient as the unit of evidence and ask which pipeline component carries signal. Across 11 frozen backbones, four pre-specified gene programmes and 285 TCGA-BRCA patients with paired slides and RNA-seq (44 cells; GroupKFold by patient, all preprocessing fitted inside the fold), ridge regression on mean-pooled embeddings predicts held-out programme scores at Spearman rho = 0.25-0.56, UNI2 strongest on all four (immune 0.556). A matched permutation null gives raw p ~ 1e-4 at 10,000 permutations for every cell; Holm-adjusted p = 0.0044. The signal is real but not uniformly morphological. Against competing models on the same patients and folds, embeddings beat tissue composition for ER/luminal, proliferation and immune (+0.280, +0.284, +0.479; p <= 0.003) but not basal, where compartment fractions alone reach 0.469 against the embedding's 0.493 (p = 0.77). Fifty-four interpretable cell-count features come within 0.043-0.085 on every programme. The geometric machinery contributes nothing measurable, and we identify why: the geodesic graph selects neighbours by Euclidean nearest-neighbour search and only reweights edges already chosen, so the topology is Euclidean by construction (Riemannian minus Euclidean = +0.0010, 95% CI [-0.0007, +0.0029]). Applied consistently the geometry is worse (-0.0117). Ridge regression beats the graph-and-metric decoder by +0.097 (CI [+0.069, +0.127]). The driver-count metric common in this literature is near-uninformative here: 91.8% of random six-gene panels recover >=5/6 drivers.
Authors: Yanbin Hu, Jin Cui, Jun Ye, Jiepeng Zhou, Jiangcheng Song, Boran Zhao, Pengju Ren
Abstract: 3D scene understanding requires reasoning about entity existence, spatial layout, and object relations, yet RGB images alone often provide insufficient 3D cues. Existing 3D-VLMs commonly rely on depth or 3D-position-aware inputs at inference time, introducing additional acquisition, reconstruction, or annotation costs that limit RGB-only deployment. We therefore study how training-time 3D evidence can be converted into spatial reasoning capabilities retained under RGB-only inference. We propose a privileged-evidence distillation framework that constructs a distillable teacher through a unified evidence interface and controlled residual injection, and transfers its knowledge to a deployable student receiving only RGB images and questions through logit and structured representation distillation. To avoid imitating teacher signals unsupported by RGB, we further introduce evidence-sensitivity-guided distillation, which uses corrupted evidence to identify highly evidence-dependent targets and down-weight their supervision. We also define a recoverability decomposition based on the matched baseline, teacher, and student, separating privileged gains into RGB-recoverable improvements and residual teacher advantages. Across four benchmarks, the teacher achieves the best result on 7 of 11 reported metrics among the compared methods. The RGB-only student outperforms its matched baseline on all 11 metrics, including gains of 10.4 ScanQA CIDEr and 19.1 Scan2Cap CIDEr@0.5, without additional inference-time inputs. These results validate the effectiveness of training-time privileged 3D evidence distillation for both teacher performance and deployable RGB-only spatial reasoning. Separately, our matched baseline-teacher-student analysis characterizes privileged-gain transfer across evidence types and spatial skills.
Authors: Saleh Sakib Ahmed, M. Sohel Rahman
Abstract: Rapid estimation of impacted structures - critical for conflict-zone humanitarian response - is frequently hindered by post-strike satellite data embargoes and imagery blackouts. We bypass this operational bottleneck by reframing impacted building mapping as a zero-shot geometric projection task on archival, pre-strike maps. Using coordinate and incident text from LiveUAMap and ArcGIS, Large Language Models extract weapon payloads (W) to project kinetic blast perimeters via Hopkinson-Cranz scaling (R_base = Z * W^(1/3)). To count exposed structures within these zones without post-strike imagery, we introduce two technical innovations: Adaptive Field-of-View to eliminate resolution (zoom) bias in 2D segmentation (SAMGeo), and 2.5D pseudo-height depth maps combined with segmentation masks to help Large Vision-Language Models (LVLMs) resolve overlapping, dense rooftops. Evaluated on 2026 Middle East conflict data, depth-augmented LVLMs dramatically outperform traditional segmentation in congested urban centers. This establishes a powerful hybrid paradigm for zero-shot crisis mapping: ultra-fast 2D segmentation for sparse rural zones, and depth-augmented LVLMs for dense urban environments.
Authors: Fabian Drexel, Marlene Fritzsche, Era Stambollxhiu, Miriam Kumpf, Lena Schmitzer, Lea Schumann, Jannik Kahmann, Friedrich Puttkammer, Johannes Moll, Jannik L\"ubberstedt, Zeineb Ben Chaaben, Anirudh Narayanan, Cosmin I. Bercea, Sebastian Ziegelmayer, Marcus R. Makowski, Daniel Rueckert, Lisa C. Adams, Keno K. Bressem
Abstract: Vision-language pretraining learns rich medical image representations from radiology reports, but previous model variants commonly operate within a single shared embedding space, so concept-level structure and interpretability must be recovered post hoc, limiting model transparency and, hence, clinical utility. We introduce RadPRISM, which makes a clinician-defined radiology schema a designated stratification axis: an on-premise large language model extracts per-concept text spans from free-text reports, and each clinical concept is aligned in its own dedicated visual subspace, turning concept stratification into direct, top-level alignment supervision. Instantiated on chest radiographs with a 19-concept schema over $203{,}602$ examinations from an internal multi-year archive, RadPRISM improved internal dataset zero-shot classification from $0.717$ (95% CI, $0.710-0.723$) to $0.868$ (95% CI, $0.863-0.872$) macro AUROC over a matched global-alignment baseline, performed on par with the purpose-built CARZero reference in external zero-shot classification while substantially outperforming it (up to 4.3-fold) in pointing-game visual grounding. In addition, a radiologist reader study demonstrated concept-stratified retrieval ability ($0.78$ macro retrieval correctness rate within rank 3), surfacing disentangled descriptive findings that report-level retrieval and fixed-label vocabularies cannot express. RadPRISM yields discriminative, spatially faithful, natively concept-stratified representations shaped by and transparently inspectable by clinicians.
Authors: Nawazish Khan, Sanjay Bhandari, Sarang Joshi, Alzbeta Novotna, Tiffany Jeong, Loretta Bowman, Michael Hernandez, Tobi Somorin, Viraj Govani, Jesse Glodstein, Shireen Elhabian
Abstract: Correspondence-based statistical shape modeling (SSM) is vital for population-level morphometric analysis, but conventional pipelines assume clean, fully registered surfaces. Real-world clinical photogrammetry scans are often noisy, partial, and cluttered, hindering the adoption of radiation-free surface imaging as a safe alternative to computed tomography (CT) for infant craniosynostosis. We present SCALP (Semi-supervised Correspondence via lAndmark Localization and sPectral warping), a two-stage framework that constructs consistent shape models directly from raw, imperfect surface scans. First, a semi-supervised Point Transformer leverages a small expert-annotated dataset alongside a large unlabeled cohort to accurately localize craniofacial landmarks with minimal annotation overhead. Second, these landmarks anchor a Laplace--Beltrami spectral deformation of an anatomical template, generating dense correspondences while naturally isolating the cranium from peripheral scanning clutter without manual preprocessing. Experiments on infant photogrammetry scans demonstrate that SCALP consistently outperforms state-of-the-art unsupervised point-cloud approaches, offering a clinically practical pathway toward objective, radiation-free head shape analysis.
Authors: Boyang Li
Abstract: 3D Gaussian Splatting (3DGS) gives high-quality novel-view synthesis, but its adaptive radiance primitives are not directly usable as structured assets: opacity is not an additive area measure, refinement can change the induced geometry, and watertight mesh extraction can hallucinate collision surfaces in unobserved regions. We introduce Manifold-GS, a certified hybrid asset layer for Gaussian scenes. The method separates appearance opacity from geometric quadrature mass, represents surface-like Gaussians as a discrete unoriented varifold, and exports only confidence-certified open surface patches while retaining uncertified content as residual splats. It provides refinement-conservative mass transport, local realizability diagnostics, source-preserving patch bindings, and conservative collision candidates. On three DTU scenes, a frozen asset benchmark shows zero patch-defined edit leakage, texture round-trip PSNR of 30.1/35.3/33.7 dB, and lower collision floater area than official 2DGS meshes on all scenes, with large gaps on two scenes. The result is a precision-coverage tradeoff rather than a universal reconstruction claim. External-region annotations, phantom-collision probes, and 5k-face simplification further support the certified asset interpretation, while RGB-only experiments show that local realizability is not sufficient for sparse-view surface identifiability.
Authors: Guoliang You, Haifan Gong, Xiaomeng Chu
Abstract: Volumetric CT vision-language pretraining learns 3D representations from scan-report pairs, but global and anatomy-aware objectives supervise only correspondence: they establish what is present and leave how much unconstrained. Nothing separates a mild from an extensive case of the same finding along a consistent direction, so the graded burden language in reports collapses into a present/absent signal. Longitudinal supervision would supply this order, but patient-matched CT pairs are scarce at scale; cross-sectional cohorts already encode weak burden cues across different patients. We introduce Spectrum, an anatomy-conditioned framework that represents each study at whole-study and organ scopes. For each organ-mapped pathology, a rule-based scorer mines confidence-filtered lower-to-higher pairs of different patients, and Burden-Direction Alignment (BDA) aligns the pathology-conditioned image delta with the report delta at each scope, separating that direction from its reverse. Because the endpoints are different people, a target-conditioned aligner first makes them comparable, so the delta reflects burden rather than between-patient variation. BDA further separates the selected direction from its reverse, anchors it to the observed higher-burden endpoint, and enforces consistency across ordered triplets. Since every pair is drawn within a single pathology, BDA is designed to constrain intra-class structure that image-report contrast alone never touches. Spectrum attains 85.6 zero-shot AUROC on CT-RATE and 72.7 on external RAD-ChestCT, with consistent gains in linear probing and retrieval. Weak cross-patient order is thus a scalable complement to anatomy-aware correspondence, yielding burden-aware CT representations without longitudinal data.
Authors: Shih-Chin Yang, Yanming Xiu, Hanting Ye, Qi Chen, Elias Rotondo, Maria Gorlatova
Abstract: Surgical augmented reality (AR) can provide contextual guidance by overlaying virtual annotations, tool cues, and procedural information onto the surgical workspace. However, the virtual content may obstruct task-relevant real-world information, such as surgical instruments, and interfere with users' perception during time-sensitive surgical tasks. In this paper, we investigate visual obstruction detection for surgical AR and present a latency-aware pipeline that combines vision-language model (VLM)-based surgical-object recognition with segmentation-based obstruction reasoning. To reduce inference overhead, the system adopts a cascaded small-to-large VLM architecture with segmentation-guided early exiting and attention-based visual token pruning. The small VLM handles easy frames when its key-object prediction is supported by segmentation consistency, while difficult frames are forwarded to a large VLM with pruned visual tokens. We construct a pseudo-AR surgical obstruction detection benchmark by overlaying virtual content onto surgical-tool images and labeling whether the virtual content obstructs task-relevant instruments. Evaluation results show that the proposed system achieves 87.43% obstruction detection accuracy with an average end-to-end latency of 479 ms, reducing latency by 62.90% compared with a cloud large-model baseline. These results demonstrate the feasibility of latency-aware obstruction detection for surgical AR and motivate future work on dynamic surgical videos, multi-object scenes, and clinically grounded AR guidance content.
Authors: Meibo Hu, Guohao Sun, Annemarie D. Ross, Sheng Li, Zhiqiang Tao
Abstract: Vision-language models (VLMs) have emerged as a powerful framework for multimodal video understanding. However, they remain limited in the sign language translation task, where we identify a key failure mode of existing VLMbased translators: poor spatial-temporal visual grounding. In particular, we find that standard next-token cross-entropy does not directly provide signal for where and when the model should attend, causing models to overlook sign-relevant regions and frames. To address this challenge, we propose AttnSign, a VLM-based spatial-temporal attention steering framework for sign language translation. AttnSign first introduces spatial attention supervision for sign-relevant regions, such as face and hands, in each frame; then develops an RL-based motion-cadence steering method that encourages the model to explore and focus on sign-level keyframes. Experimental results on How2Sign and OpenASL benchmarks show that our proposed AttnSign consistently outperforms existing methods.
Authors: Meibo Hu, Jiamian Wang, Pichao Wang, Zhiqiang Tao
Abstract: Vision-language models (VLMs) have recently emerged as a promising paradigm for end-to-end autonomous driving, enabling agents to map multimodal inputs and high-level navigation instructions directly to executable trajectories. However, in practice, these models exhibit a persistent command-following gap: predicted trajectories often show weak sensitivity to navigation commands, resulting in incorrect behavior at critical decision points. We identify this issue as a form of conditional policy collapse, where regression-based training under multimodal trajectory distributions encourages the model to rely on dominant visual priors while marginalizing the language-conditioned signal. To address this issue, we introduce a principled formulation of classifier-free guidance (CFG) for regression-based vision-language driving. We show that CFG can be interpreted as isolating the instruction-induced residual in the action space by contrasting conditional and unconditional predictions, thereby explicitly amplifying the effect of the navigation command at inference time. However, a standard two-pass CFG introduces prohibitive latency for real-time control and produces noisy instance-level guidance directions. Building on a mean-shift interpretation of CFG, we propose Latent-Centroid Steering (LCS), a single-pass guidance mechanism that replaces instance-level residuals with class-level latent shifts. By projecting conditional representations toward precomputed command-specific centroids, LCS performs class-level latent steering based on cluster geometry that is both more stable and computationally efficient. We demonstrate that LCS reduces inference latency by approximately 50% while achieving stronger command adherence and improved driving performance on both closed-loop (Bench2Drive) and open-loop (nuScenes) benchmarks. Code will be released.
Authors: Guoliang You, Haifan Gong, Xiaomeng Chu
Abstract: Learning transferable representations from CT-report pairs requires combining whole-volume context with anatomy-specific evidence. Existing methods typically emphasize either global CT-report alignment or fine-grained anatomy-level correspondence. Global alignment preserves broad study context but leaves the contribution of localized evidence implicit, whereas anatomy-level alignment explicitly grounds local findings but does not specify how independently represented evidence should interact, acquire study-level meaning, and contribute to a global CT representation. To address this gap, we propose SCOPE (Semantic Calibration Of comPosed Evidence), a framework for semantically calibrated evidence composition in CT vision-language learning. Under organ-specific report supervision, mask-guided queries with fixed anatomical identities extract context-aware organ evidence from shared, uncropped volumetric features, while an unrestricted global query retains access to whole-volume context. The global query then drives Local-Global Coupling to compose the organ evidence into a unified evidence representation. The composed evidence is subsequently calibrated using the diagnostic summary, providing study-level semantic supervision beyond local organ descriptions, and is finally integrated as a controlled residual into a context-preserving whole-volume representation aligned with the complete report. This progressive pathway connects localized evidence with study-level semantics without reducing the CT representation to a predefined set of organs. On CT-RATE and RadChestCT, SCOPE achieves macro AUCs of 85.0 and 72.2, respectively, outperforming the previous SOTA by 7.2 and 4.2, while also yielding substantial gains in linear probing and cross-modal retrieval. These results demonstrate the effectiveness of semantically calibrated evidence composition.
Authors: Andrea Filiberto Lucas, Mark Bugeja, Carl James Debono, Dylan Seychell
Abstract: Automated visual monitoring of urban environments is a growing Computer Vision research area, but municipal solid waste detection remains under-represented in dedicated benchmark resources. Existing waste-related datasets predominantly address individual litter detection, aerial imagery, or image-level classification, and none simultaneously provide street-level imagery, instance-level localization, and categorization of domestic waste streams within a structured municipal collection context. This paper introduces the Maltese Domestic Waste Dataset (MDWD), a street-level benchmark comprising 3,697 high-resolution images and 11,461 manually annotated instances across five domestic waste categories representative of Malta's municipal collection system. The dataset captures substantial variation in location, illumination, object scale, occlusion, and urban context. To establish reproducible baselines, a cross-architecture benchmark is conducted across multiple generations of the YOLO family and a transformer-based detector. On the test set, RF-DETR-M achieves the strongest overall performance with an mAP50 of 94.49% and an F1-score of 93.56%, whilst smaller-capacity variants maintain competitive accuracy at substantially reduced parameter counts. These results indicate that MDWD supports effective training across both compact real-time detectors and transformer-based models, establishing a benchmark for future research in vision-based municipal waste monitoring.
Authors: Kamil Ksi\k{a}\.zek, Piotr Suszy\'nski, Micha{\l} Jan W{\l}odarczyk, Jacek Tabor, Przemys{\l}aw Biecek
Abstract: Vision foundation models, such as DINOv2, learn highly expressive representations but rely on massive, opaque architectures that demand substantial computational power and memory. To provide an interpretable-guided and efficient solution to this issue, we first propose a spectral analysis and new visualization technique for individual attention heads based on the Laplacian eigenvectors of their attention maps. Building upon recent observations regarding the block structure of Vision Transformers, we perform semantic clustering of attention heads and identify functional redundancies. Leveraging these insights, we introduce SAPER (Soft Attention PrunER), an end-to-end differentiable pruning framework based on the LapSum Soft Top-K approach. Extensive experiments on ImageNet-1K demonstrate that SAPER achieves a highly favorable accuracy-efficiency trade-off, outperforming the competitive RAPTOR baseline in FLOPs reduction while preserving strong classification performance.
Authors: Renjie Liang, Zijian Xu, Jinqian Pan, Chengkun Sun, Zhengkang Fan, Shawn Li, You Qin, Mei Liu, Jie Xu
Abstract: A 3D CT scan entering a vision-language model produces a long sequence of visual tokens, often thousands to tens of thousands per volume, and this sequence must be compressed before a language model can consume it. Token compression is well studied in general vision, but little of it targets 3D CT specifically. A common baseline is grid average, which pools regular grid cells and can blend distinct anatomy, lesion, and air into one token. We present \textbf{ORCA} (ORgan-Centroid Aggregation), a token compressor for 3D CT. It merges adjacent tokens with organ guidance and adds a sinusoidal encoding of each region's centroid to preserve spatial layout. This preserves the anatomical information a downstream model needs. ORCA is training-free and plug-and-play, producing an adjustable token set without any model change or text query. We evaluate it across two datasets (CT-RATE and Merlin) and five encoders. The evaluation spans two task types: attribute prediction over five families (size, density, location, texture, and disease) and text generation (visual question answering and report generation). At matched token budgets, ORCA improves consistently over existing compression methods. It shrinks the visual context $64\times$ and its KV-cache $50\times$, and is $31\times$ faster to process each volume. Code released at https://github.com/renjie-liang/ORCA-3DCT.
Authors: Boyi Li, Yifan Shen, Houze Yang, Xu Cao, Guojun Yun, Li Gao, Turong Chen, Long Xu, Jianguo Cao, Meihuan Huang
Abstract: The First AI Children Challenge aims to advance real-world applications of computer vision and AI in child healthcare, child education, and pediatrics. The 2026 CV4CHL edition featured the first track in this domain: Children Gait Visual Analysis. The main goal of Children Gait Visual Analysis is the fine-grained analysis of children's gait behaviors from keypoint sequences. This is still a big challenge for human action recognition. Experienced medical doctors can distinguish these subtle nuances, but none of the people test AI models in this domain. To bridge this gap, we introduce thousands of 2D children keypoint sequences walking around videos across various age groups of children (3-16 years old). There is a significant opportunity for batch analysis of these videos to provide clinically relevant insights into medical diagnosis. The Challenge will be launched with two problem tracks: Edinburgh Visual Gait Score (EVGS) Scoring and Classification of Gait Patterns in Bilateral Spastic Cerebral Palsy. Each track is chosen in consultation with board-certified pediatricians based on the value of potential solutions. With the first available dataset for such tasks and ground truth for each track, the challenge enabled participants to evaluate their solutions. Final rankings will be revealed after the competition concludes, fostering reproducibility and mitigating overfitting.
Authors: Simiao Sun, Kenneth Ng, Lynn Lee, Astrid Harth, Asami Odate, Aggelos Katsaggelos, Manuel Ballester Matito, Nicholas Eastaugh, Marc Walton
Abstract: Optical microscopy of particle and fiber dispersions involves interpreting subtle visual cues influenced by specimen morphology, chemical composition, magnification, and illumination conditions. We introduce an artificial intelligence (AI) distillation framework that extracts semantically rich image embeddings from microscopy images using semantic anchors. A multimodal teacher combines each image's visual embedding with three text embeddings representing illumination modality, magnification, and specimen identity and morphology. Generated by LongCLIP's extended-context text encoder, this yields a 2304-dimensional block-structured teacher vector whose component blocks remain physically interpretable throughout training and inference. A student vision transformer (ViT) with a multi-layer perceptron (MLP) decoder is trained to reconstruct this teacher vector from the image alone, minimizing a mean absolute error (L1) loss that enforces coordinate-level fidelity to the teacher's block structure. A cross-entropy term over pseudo-classes derived from HDBSCAN clustering of the teacher embedding space acts as a collapse-prevention regularizer, enforcing inter-cluster separation without requiring contrastive negative mining. At inference, the student operates on image input alone, producing compact embeddings that recover the full semantic content of the teacher vector. The framework achieves approximately 80% pseudo-class validation accuracy and 75% Recall@1 on fine-grained specimen description labels under leave-one-out nearest-neighbor retrieval. These results demonstrate that semantic anchoring enables a vision-only student to acquire richer and more interpretable representations than image-only training, with direct applicability to retrieval, classification, and exploratory analysis of heterogeneous particle and fiber dispersions.
Authors: Yifan Shen, Boyi Li, Meihuan Huang, Yuanzhe Liu, Xu Cao, Jinyang Jin, Zhengyuan Li, Anglin Liu, Junho Kim, Jingyuan Zhu, Lan Fangzhou, Jianguo Cao, Jintai Chen, Ismini Lourentzou, James Matthew Rehg
Abstract: We introduce a new problem domain for human action recognition: the fine-grained analysis of children's gait behaviors from standard RGB video. We specifically target the ambulatory patterns of children aged 3-17 years. Such behaviors arise naturally in the diagnosis and treatment of several critical developmental and neuromuscular disorders, such as cerebral palsy and hemiplegia. Despite their clinical value, current 3D sensor-based gait analysis systems are expensive, intrusive, and often impractical for young subjects. To address this, we introduce a new dataset comprising over 1,100 high-frame-rate (60 FPS) video sequences from 110 subjects, accompanied by synchronized, anonymized pose sequences. In each session, the child performs a 5-second "walk-around" task, capturing the gait cycle from multiple viewpoints. Crucially, we demonstrate that current state-of-the-art approaches, including gait foundation models and Multimodal Large Language Models (MLLMs), fail to effectively resolve these clinical nuances. We identify the key technical challenges in analyzing these erratic and subtle motor patterns and describe a unified end-to-end framework for decoding fundamental components of pediatric gait. Through comprehensive experimental results, we demonstrate the potential of this dataset to drive novel research questions and establish a rigorous baseline for automated child gait assessment.
Authors: Liangjing Shao, Beilei Cui, Yiming Huang, Changjing Liu, Hongliang Ren
Abstract: Depth estimation is a significant task for 3D perception in endoscopic surgeries. However, illumination interference and feature diversity in various endoscopic scenes are still challenges for generalizable depth estimation and ego-motion estimation. Based on this, a novel self-supervised framework, EndoMINI, is proposed for depth estimation in endoscopic scenes. Specifically, mixture of low-rank experts (MiLoRE) is proposed to perform parameter-efficient fine-tuning, which can also boost the model adaptation to scenes with different characteristics. Meanwhile, an intrinsic image alignment (IIA) is introduced into the training loss to alleviate the influence of light reflectance in endoscopy with a novel intrinsic image decomposition network. The proposed method is evaluated on SCARED datasets for supervised depth estimation, and two endoscopic datasets, Hamlyn and SERV-CT, for zero-shot depth estimation, compared with state-of-the-art works as well. The experimental results demonstrate outstanding performance of the proposed model and the effects of the main contributions.
Authors: Zhishan Zou
Abstract: Recent image generators can synthesize convincing human-centric images, yet producing a useful collection remains different from producing a single successful image. A human-centric dataset must cover varied people and contexts, avoid implausible attribute combinations, preserve an everyday photographic character, and expose quality-control decisions at scale. We present Poplar, a reproducible Specify--Render--Inspect pipeline for human-centric image dataset synthesis. Specify samples structured attributes under commonsense constraints and verbalizes them as photography-oriented prompts. Render uses a realism-adapted image generator across composition-aware aspect ratios and retries obvious technical failures. Inspect applies a single structured vision--language review to each candidate, preserving the original prompt while rejecting intrinsic image defects or material prompt mismatches. Using Poplar, we construct Poplar-9K: 9,401 curated human-centric image--text pairs retained from 11,765 reviewed candidates (79.9\% acceptance). We release the dataset together with the pipeline, configurations, immutable generation prompts, and auditable inspection records as a compact resource for building customizable human-centric collections.
Authors: Yibo Wan, Jinyu Cai, Seekiong-Ng
Abstract: Medical anomaly detection identifies abnormal images and localizes lesions under scarce supervision while generalizing across organs and modalities. Existing CLIP-based methods reduce annotation requirements through vision--language alignment, but their normal and abnormal references, whether text prompts or learned visual tokens, remain fixed across test images. Such static references may not transfer reliably to unseen targets in a cross-domain medical imaging scenario. To address this, we propose ReCAP, a language-free framework that replaces static anchors with input-conditioned visual prototypes. ReCAP re-centers separated normal and abnormal prototypes for each image through a bounded gated modulation, enabling query-adaptive anomaly scoring while constraining context-induced prototype drift. For the few-shot setting, we introduce a non-parametric normal-reference memory to preserve instance-level target-domain variation and complement the conditional prototype branch. Across six medical benchmarks, ReCAP achieves the best image-level AUROC on all zero-shot and 23 of 24 few-shot settings, and the best zero-shot pixel-level AUROC on all three segmentation datasets. Particularly, it reduces inference latency by over 70% compared to the fastest baseline, without text prompts or test-time gradient updates.
Authors: Yibo Wan, Jinyu Cai, Yunhe Zhang, Yi Bin, See-kiong Ng
Abstract: Unsupervised medical anomaly detection learns normal anatomical patterns from healthy training images and identifies deviations at test time. Reconstruction-based and diffusion-based methods commonly use the difference between an input image and its reconstruction as anomaly evidence. However, this residual can be ambiguous. Expressive models may preserve pathological structures, while benign anatomical variation, imaging noise, and acquisition differences may also produce large reconstruction errors. We propose discriminative mask-guided diffusion (DMD), a medical anomaly detection framework that complements residual-based localization with reconstruction-shift discrimination. DMD first learns a compact quantized latent representation of normal images. Localized masks then perturb selected latent regions, and a latent diffusion model reconstructs the perturbed representations. The resulting reconstructions are paired with their original normal images to define a self-supervised classification task. At inference, the classifier provides a learned image-level anomaly score, while the residual between the input and its diffusion-based reconstruction yields a pixel-level anomaly map. Experiments on five datasets spanning brain MRI, breast ultrasound, and chest radiography show that DMD achieves the best overall performance among the state-of-the-art baseline methods.
Authors: Huu Phong Nguyen, Delower Hossain, Ehsan Saghapour, Zhandos Sembay, Jake Y. Chen
Abstract: Lung cancer results in roughly 1.8 million fatalities annually worldwide, with non-small cell lung cancer (NSCLC) comprising the majority of cases. Despite advancements in treatment, survival stratification remains challenging due to intratumoral heterogeneity inadequately captured by conventional descriptors. Standard radiomic and deep learning techniques regard imaging features as independent quantities, overlooking structured interactions between tumor characteristics. We evaluate whether structured proxy features can enhance multimodal NSCLC survival prediction by augmenting pretreatment computed tomography (CT) representations, radiomics, and clinical variables with six simulation-derived features designed to capture interactions between heterogeneity and morphology. A radiomic-parameterized cellular automaton generates growth-rate and necrosis-ratio proxy features from baseline CT by using entropy and sphericity to compute low-dimensional proxy parameters. The imaging backbone is a Transformer-based Masked Autoencoder (TMAE), which was chosen after a systematic evaluation with alternative encoders within the same pipeline and provides attention-based visualizations that highlight tumor regions receiving higher model attention. On the public Lung1 cohort (n = 390), the primary four-modality fusion attained a C-index of 0.641 (iAUC 0.731, log-rank p < 0.001). The primary result compares favorably with prior multimodal results on Lung1 (C-index 0.631; iAUC 0.592 [15]) under a comparable evaluation protocol, while a separate exploratory coefficient-optimization analysis achieved a best observed C-index of 0.662 (iAUC 0.748). These results indicate that, in addition to conventional radiomic, deep, and clinical representations within the Lung1 benchmark, simulation-derived proxy features may provide complementary predictive information within this fixed Lung1 benchmark.
Authors: Masaki Yoshida, Ren Togo, Takahiro Ogawa, Miki Haseyama
Abstract: 3D Gaussian Splatting (3DGS) turns captured or generated imagery into photorealistic 3D world simulations that users can freely explore, yet these worlds remain silent. Because existing audio generation methods condition on a single image or viewpoint, their sound is tied to that observation and cannot stay consistent while a listener moves. We introduce the task of generating a spatially consistent soundscape for a given 3DGS world through auditory grounding, identifying which objects in the world should emit sound and anchoring each to a persistent 3D position, and present Scene2Sound, a training-free framework built on this grounding. From the input world alone, our pipeline selects viewpoints that jointly cover the scene, identifies sound-emitting objects with a vision-language model, and associates the multi-view detections into 3D instances through Gaussian set matching, which measures the overlap between the Gaussian sets that render each detection. Each source then receives generated audio that a standard object-based audio engine spatializes in real time at arbitrary listener poses. We further propose two spatial-consistency metrics, one testing whether rendered audio responds consistently to listener motion and one testing whether the claimed sources are supported by views held out from their placement. On a curated set of generated 3DGS worlds and on 3DGS scenes generated from real-world 360-degree captures, Scene2Sound preserves the audio quality of strong per-viewpoint baselines while remaining spatially consistent where per-viewpoint and single-panorama pipelines do not, and a user study confirms the perceptual benefit. Project page: https://masaki-lmd.github.io/scene2sound/.
Authors: Kaho Li, Pengyu Zeng, Yuqin Dai, Jun Yin, Tianjing Feng, Shuai Lu
Abstract: Architectural drawings violate the usual assumption behind multi-view reasoning: plans and sections are cuts, while elevations are facade projections, so corresponding components change appearance in ways camera motion cannot explain. We introduce CrossProjection, an anchor-grounded diagnostic of whether vision-language models preserve component identity and externalize geometry across heterogeneous architectural views. It evaluates Matching, Registration, and Geometric Grounding through categorical judgments, candidate selection, and free point, line, and region localization. Across 23 real drawing sets and 1,954 categorical conditions per model, GPT-5.5 scores 82.4%, Qwen3-VL-32B-Instruct 62.2%, and GLM-4.5V 57.2%. A matched 200-target study crosses natural and vector-text-suppressed drawings with closed-candidate and free-geometry outputs. Candidate-supported performance is often higher, but free localization remains fragile: on natural drawings, point/region PCK@.05 is 54-76% for GPT, 8-10% for Qwen, and 14-36% for GLM; line endpoint PCK@.05 is 22%, 4%, and 0%. A coordinate grid recovers some GPT point/region precision but not lines. Three architecture-trained participants reach 87.3-93.3% categorical accuracy and 76-92% GT-region hit, supporting task feasibility rather than a population-level human ceiling. Because the categorical families do not form a same-item Matching-Registration contrast and interface controls alter multiple burdens, we avoid mechanistic claims. The supported conclusion is narrower: closed-choice or marked-element success does not entail reliable explicit geometric grounding. For drawing-guided CAD/BIM systems, categorical correctness should not be treated as evidence of candidate-free spatial reliability. Reusable on-sheet anchors, fixed-denominator scoring, and hash-locked artifacts establish an audit trail for this gap.
Authors: Tongsheng Ding, Zhen Luo, Yixuan Yang, Boyu Wang, Luyang Xie, Jinyu Yang, Feng Zheng
Abstract: Accurate prediction of object trajectories during manipulation is essential for closing the perception-action loop. Progress is limited on two fronts: available datasets lack fine-grained language-to-motion annotations, and existing predictors either rely on privileged inputs such as video, depth, or CAD models, or recover motion from fully generated videos through costly, error-prone perception pipelines. We close the supervision gap with the MOVE dataset, 5,038 object-centric egocentric trajectories, each paired with a fine-grained natural-language instruction rather than a coarse verb-noun label. We further propose DreamTraj, which predicts a 6-DoF object trajectory from a single RGB image and a task instruction, requiring no video, depth, or CAD model at inference: rather than generating a video, it reads motion from the internal representations of a frozen image-to-video diffusion model at an early denoising step. A lightweight flow-matching Reader decodes query-key attention tracks and pooled hidden states into relative 6-DoF poses. To our knowledge, this is the first approach to directly decode object 6-DoF trajectories from intermediate video diffusion representations rather than generated pixels. DreamTraj sets a new state of the art on both translation and rotation against forecasters that consume multi-frame or privileged inputs, and runs 4.6x faster than generate-then-extract pipelines.
Authors: Wendi Liu, Yujie Lu, Zengxi Zhang, Haiyang Jiang, Weihang Ran, Yinqiang Zheng
Abstract: Single-photon avalanche diode (SPAD) cameras are promising for low-light and high-dynamic-range intensity imaging, but their practical use is limited by complex sensor-specific noise. Unlike time-correlated single-photon counting (TCSPC) systems, SPAD cameras record whether at least one detection occurred in each gate without photon timestamps in intensity imaging mode, making explicit noise decomposition difficult. We present a practical noise modeling and calibration framework for SPAD intensity denoising. Our forward model describes binary-frame accumulation with a Binomial observation process, models signal-independent dark noise as an exposure-dependent pure dark count term plus an exposure-independent dark-frame bias term, and incorporates pixel-wise response non-uniformity. We design a dedicated calibration procedure for the proposed model and use it to build a count-domain noise-synthesis pipeline for network training. For denoising, we further design a SPAD-specific dark-shading correction (SPAD-DSC) to remove most systematic noise before network training. We construct a real-world SPAD intensity dataset for testing. Experimental results demonstrate the superiority of the proposed noise model.
Authors: Misora Sugiyama, Toya Oyama, Hirokatsu Kataoka
Abstract: Can image-editing models discover visual rules in image space and complete problem-solving end-to-end? We tackle this question in the spirit of a human worksheet test (e.g., an IQ test), using problems that require models to read image-based instructions, recognize the problem, infer the answer, bind it to the correct destination, control output count, suppress unnecessary edits, and preserve the input and format. We introduce WISRD, a Worksheet Image-Space Rule Discovery benchmark with 11 core tasks under eight information conditions, spanning localized marking, filling, copying, counting, and no-edit suppression, together with four supplementary reasoning-stress probes for multi-step spatial manipulation, abstract pattern reasoning, logical inference, and constraint-based problem solving. We identify three key findings as follows. (i) Among the frontier image-editing models evaluated, Nano Banana Pro achieves the highest score. On the shared V0--V3 no-reference subset, the Auto-Strict proxy pass rates are 48.7% for Nano Banana Pro, 13.4\% for Qwen-Image-Edit, 11.5% for FLUX.2 Klein 4B API, 11.3% for FLUX.2 Klein 4B open-weight, and 0.0% for InstructPix2Pix. (ii) Analysis reveals that current image-editing models can partially rely on rendered in-image instructions even when the external prompt is absent or merely generic. (iii) In small supplementary diagnostics, Nano Banana Pro achieves 70.0% on 4-by-4 Sudoku and 22.9% on public RAVEN pattern-discovery items in image space.
Authors: Wendi Liu, Weichao Zeng, Weihang Ran, Yujie Lu, Yinqiang Zheng
Abstract: Optical flow remains challenging in high-speed and low-light scenes, where the limited frame rate and sensitivity of conventional cameras lead to motion blur and underexposure. Single-photon avalanche diode (SPAD) cameras offer single-photon sensitivity and extremely fine temporal sampling. However, individual slices in these high FPS binary photon streams are too sparse for dense correspondence. Temporal aggregation can provide the spatial cues required by optical flow, but accumulating photons at fixed coordinates blurs moving structures. Motion-aware aggregation can reduce this blur, yet it depends on the flow being estimated. To address this dependency, we propose QuantaFlow, the first method for dense optical flow directly from SPAD streams. Instead of constructing a fixed input representation, QuantaFlow embeds SPAD representation construction into iterative flow refinement. At each iteration, the current flow coarsely aligns the slices within the source and target sub-streams. A photon-flux transformation then constructs multi-scale representations containing intensity and structural cues, while adaptive multi-scale fusion balances photon noise and residual motion blur at each pixel. The fused representations drive a feature-warping flow update, and the refined flow guides representation construction in the next iteration. We further construct a synthetic dataset for SPAD optical-flow training and evaluation. Experiments on the synthetic dataset and real-world SPAD data demonstrate the effectiveness and generalization of QuantaFlow.
Authors: Yufei Zhang, Chenlu Zhan, Donghui Sun, Xiaoxin Chen, Hongwei Wang
Abstract: Affordance grounding aims to localize the functional region for interaction, such as the handle to grasp or the button to press, rather than the whole object. This makes it more challenging than generic visual grounding because the target region is smaller, more ambiguous, and more dependent on task context, especially for compact vision-language models (VLMs) used in embodied settings. Recent sequence-level supervision and reinforcement learning improve coordinate prediction quality, yet compact autoregressive VLMs still lack reliable affordance-aware visual focus before coordinate generation: the model can produce better coordinate tokens while its cross-modal attention remains diffuse and weakly anchored to the true affordance evidence. To address it, we propose SpatialAfford, a two-stage framework that first aligns attention to the ground-truth affordance region through Spatial Attention Alignment (SAA), then refines coordinate prediction with Spatial-Aware GRPO. By explicitly teaching the model where to look before optimizing where to ground, SpatialAfford turns affordance grounding from a purely output-constrained objective into attention-grounded spatial reasoning. Across ShareRobot-Bench, ReasonAff, and PartAfford, SpatialAfford consistently improves affordance grounding, with a compact 4B model outperforming stronger 7B+ baselines.
Authors: Kai Geissler, Laurens M\"uller-Groh, Hans Meine
Abstract: Object detection and segmentation in three-dimensional medical images is a very active area of research. However, most proposed deep learning models carry a high computational cost, and only few aim to be broadly applicable, achieve high detection performance, and remain fast to execute on resource-constrained hardware. To address this gap, we present RadYOLO, a 3D extension of YOLO11 tailored to medical images. We compare it with nnU-Net and nnDetection on five datasets comprising CT and MRI data with varying object sizes and prevalence. RadYOLO's detection performance surpasses that of nnDetection on four of five datasets and is comparable on one. Compared to nnU-Net, RadYOLO performs better on lesion detection tasks, while nnU-Net excels at detecting large organs when precise localization is required. When rough object localization is sufficient, RadYOLO matches or outperforms nnU-Net on all five datasets. Regarding inference time, RadYOLO is 8-46x faster than nnU-Net on a GPU. Compared to nnDetection the speedup is even higher. When executed on a CPU, RadYOLO's inference runs within seconds (still faster than nnU-Net on a GPU) offering a significant advantage for clinical and edge-device deployment. RadYOLO repository: https://github.com/FraunhoferMEVIS/RadYOLO
Authors: Ling Ren, Chao Deng, Ziming Wang, Yuecong Xu, Kai Zheng
Abstract: Reliable pelvic bone segmentation (PBS) from CT is essential for robot-assisted pelvic trauma surgery, yet deploying a source-trained model to a new hospital suffers from severe performance degradation due to cross-center domain shifts. While test-time adaptation (TTA) enables online model adaptation without accessing source data, existing methods show limited effectiveness for PBS, facing challenges including boundary degradation, anatomical inconsistency under domain shifts, and voxel-level class imbalance. To address these challenges, we propose a novel closed-loop dynamic Reliability-Guided TTA framework (ReGA) for PBS. Specifically, we introduce a pseudo-label reliability criterion termed Segmentation Inference Consistency Evaluation (SICE), which jointly measures region overlap and boundary deviation via dropout-based ensemble predictions. Based on SICE, a trust-weighted refinement module adaptively updates features to mitigate boundary errors in pseudo-labels. Furthermore, a confidence-weighted region-level contrastive learning strategy is proposed to enforce anatomical consistency. Finally, ReGA follows the teacher-student (TS) scheme to alleviate voxel-level class imbalance. Experiments on three heterogeneous 3D pelvic CT datasets demonstrate that ReGA consistently outperforms state-of-the-art TTA methods, enabling effective adaptation of the source-trained PBS model to unseen clinical domains. The code is available at https://github.com/Ren-ling/ReGA.
Authors: Yiwen Wang, Yuyang Deng, Yihao Long, Xi Zhao
Abstract: 3D visual grounding aims to localize the target object in a 3D scene from a natural language query, requiring both fine-grained semantic understanding and viewpoint-dependent spatial reasoning. Existing methods typically formulate semantic understanding as an auxiliary closed-set object classification task and rely on multi-view feature aggregation for viewpoint reasoning, limiting semantic generalization and weakening viewpoint-specific evidence. We observe that vision-language models naturally provide complementary capabilities through open-vocabulary semantic understanding and global scene perception. Based on this insight, we propose GuideGround, a VLM-guided framework that complements rather than replaces task-specific grounding models by leveraging VLMs for semantic enhancement and viewpoint-specific hypothesis verification. Specifically, we replace auxiliary closed-set object classification with VLM-generated object semantic descriptions to enhance semantic understanding. Meanwhile, instead of directly aggregating multi-view representations, we preserve viewpoint-specific grounding hypotheses through per-view grounding and explicitly verify them using VLMs across candidate viewpoints. Extensive experiments on the ReferIt3D benchmark demonstrate that GuideGround consistently outperforms previous state-of-the-art methods. Comprehensive ablation studies further confirm the effectiveness of both the proposed semantic understanding and viewpoint reasoning strategies.
Authors: Axi Niu (School of Computer Science, Northwestern Polytechnical University, Xi'an, China), Jieheng Li (School of Computer Science, Northwestern Polytechnical University, Xi'an, China), Kang Zhang (School of Electrical Engineering, KAIST, Daejeon, Republic of Korea), Qingsen Yan (School of Computer Science, Northwestern Polytechnical University, Xi'an, China), Jinqiu Sun (School of Aeronautics and Astronautics, Northwestern Polytechnical University, Xi'an, China), Yanning Zhang (School of Computer Science, Northwestern Polytechnical University, Xi'an, China)
Abstract: Infrared-visible image fusion under realistic degradation scenarios is a challenging task, as degradations not only cause a loss of reliable modality-specific information in observed images but also hinder the fusion process. Recent studies indicate that text can provide prior information about degradation characteristics, complementing the limited evidence available from corrupted input images and facilitating fusion. However, existing methods typically inject fixed global text representations into visual features, making it difficult for textual guidance to adapt to spatially varying degradations, local structures, and thermal saliency. To this end, we propose TGFusion, a text-guided latent-space flow matching framework that unifies degradation suppression and cross-modal fusion. TGFusion encodes task, degradation, and generation cues into structured prompts. To fully exploit these priors, we design a Prompt-conditioned Multi-stream Joint Flow Transformer that represents text as an independent semantic stream alongside fusion, visible, and infrared streams. Joint attention enables token-level bidirectional interaction and layer-wise updating among semantic and visual representations, allowing degradation semantics to dynamically guide reliable information selection and fusion latent generation. Extensive experiments on public benchmarks and complex degradation scenarios demonstrate that TGFusion achieves superior or competitive performance in perceptual quality, image naturalness, structural-detail preservation, and infrared-saliency retention, while remaining robust across diverse single and compound degradations.
Authors: Yunhao Wang, Binghong Wu, Zhenyu Huang, Jiacheng Shi, Shuo Huang, Tinghao Yu, Feng Zhang
Abstract: Reinforcement learning (RL) for document parsing often relies on reference-based rewards rooted in edit distance (e.g., tree edit distance), yet it remains hard to optimize in the high-accuracy regime because such rewards become weakly discriminative: near-correct outputs receive very similar scores, providing limited learning signal for hard cases. We propose Step-Aware Annealing (SAA), a plug-and-play reward sharpening mechanism that progressively increases reward curvature during training, amplifying subtle quality differences among high-scoring samples while preserving stability in early learning. Built on SAA, we introduce DocPO, a document policy optimization framework with element-specific, reference-based rewards anchored by edit-distance signals: normalized string edit distance (NED) for text, tree edit distance similarity (TEDS) for tables, and a hybrid Rubric+edit reward for formulas. Experiments on OmniDocBench and DocElemHard show that SAA consistently improves GRPO-style RL across document elements over non-annealed rewards, without requiring additional human supervision for reward construction.
Authors: Hongjie Wu, Yiping Xie, Jiancheng Lv
Abstract: Latent Flow Models have revolutionized compressed-space image synthesis, yet their application to high-fidelity inverse problems remains bottlenecked. In this paper, we trace this dilemma to a fundamental geometric limitation of pre-trained autoencoders, which we term \emph{First-Order Manifold Blindness}. Severe decoder compression (e.g., retaining only $\sim\!2\%$ of the original degrees of freedom) produces a rank-deficient Jacobian, rendering high-frequency measurement residuals in its orthogonal complement invisible to latent gradients even when the decoder can represent the target image. To overcome this bottleneck, we propose Hybrid-Domain Posterior Sampling (HDPS), a decoupled inference framework that disentangles physical measurement consistency from semantic prior modeling. HDPS diverges into the pixel space, leveraging Langevin dynamics to absorb precise orthogonal measurement gradients, and subsequently projects these structural corrections back onto the generative manifold. An optimization-based latent alignment is introduced to filter pixel-space artifacts while avoiding the semantic drift of direct encoding. Extensive experiments on diverse inverse problems demonstrate that HDPS establishes a new state-of-the-art, successfully recovering the high-frequency structural precision that latent-only solvers inherently discard. The code is available at \href{https://github.com/74587887/HDPS}{https://github.com/74587887/HDPS}.
URLs: https://github.com/74587887/HDPS, https://github.com/74587887/HDPS
Authors: An Lanji, Dawei Liu, Jin Li, Haoran Xu, Mei Chen, Yu Tian
Abstract: Tool-integrated vision-language agents have made remarkable progress on compositional and multi-step visual reasoning. Yet their outputs frequently exhibit unfaithfulness: the stated reasoning path diverges from the computation that actually produced the answer, undermining reliability in safety-critical applications. We present DiffuseAgent-MI, a self-evolving agent whose perceptual grounding is governed by a KL-minimal energy model over feature units, providing a distributional view of visual mechanistic interpretability. The agent learns an energy landscape that softly constrains generated samples to lie near the native prior conditioned on the chosen interpretable unit, closing the gap between the explanation and the internal representation. A verifier then supplies trajectory-level faithfulness rewards, and a repair branch re-conditions the energy when the verifier flags an unfaithful step. On GeoQA, SciVis, VQA-v2 and an in-house multimodal reasoning set, DiffuseAgent-MI improves accuracy by up to 5.1 points over prior self-evolving agents while more than doubling mutual-information faithfulness and human-interpretability agreement. Our analysis shows the energy term and the verifier are complementary: the former guarantees distributional faithfulness, the latter trajectory-level faithfulness, and only their combination closes both gaps.
Authors: Sigrid Vila-Bagaria, Mar Teixid\'o, Miquel Pi\~nol, Felip Vilardell, Robert Montal, Veronica Vilaplana
Abstract: Identifying the ``Inflamed'' immunophenotype in Gastric Adenocarcinoma predicts immunotherapy response but requires an expensive 10-gene RNA signature. While deep learning on standard H\&E slides offers a scalable alternative, conventional binary classifiers oversimplify continuous RNA data and introduce label noise. To resolve this, we propose VITA (VIrtual Transcriptomic Approximation). By aligning H\&E and RNA into a joint latent space during training, VITA requires only standard H\&E at inference to retrieve morphologically similar historical cases and approximate the continuous RNA signature. Achieving 0.72 classification accuracy and a 0.66 Spearman correlation, VITA provides a cost-effective ``virtual transcriptomics'' pre-screening tool that preserves the continuous phenotypic spectrum without requiring genomic sequencing.
Authors: Pu Cao, Qingye Kong, Xuedan Yin, Xuekun Zhao, Rupeng Yan, Qing Song, Yao Zhang, Lu Yang
Abstract: Recent image-generation models and multimodal agents can produce high-quality visuals for increasingly complex visual communication tasks. Yet their raster outputs remain difficult to use directly because meaningful content and relationships are flattened into pixels, preventing users from inspecting, modifying, rearranging, or reusing individual components. We formulate image-to-editable reconstruction, which recovers a structured, directly manipulable artifact from a raster image while preserving its visual and semantic content. The central challenge is to jointly satisfy Fidelity and Editability, which often trade off in practice. To study this task, we introduce DrawAI, comprising an agentic benchmark, DrawAI-Bench, and a reconstruction workflow, DrawAI-Flow. DrawAI-Bench spans scientific figures, presentation slides, posters, and diagrams, combining real and AI-generated images to reflect practical visual-creation scenarios. It evaluates Fidelity and Editability through a hybrid protocol of 39 criteria: deterministic rule-based metrics measure properties with direct correspondences, while asset-specific vision-language rubrics capture semantic and perceptual qualities for which exact matching is misleading. Besides, we propose DrawAI-Flow, a two-stage agentic workflow in which a Parser Agent turns extracted elements evidence into an explicit reconstruction plan, and a Reconstruction Agent realizes the plan as executable graphics code through an iterative code-render-validate-revise loop. On DrawAI-Bench, we systematically evaluate thirteen models across five agent harnesses to study the effects of model capability, harness choice, and workflow design. The results show that reconstruction quality and costs vary substantially across model-harness configurations, while DrawAI-Flow consistently improves editable structure.
Authors: Yiqian Zhang, Zheyuan Gu, Xiangzhao Hao, Zefeng Zhang, Jingjia Mao, Jiahao Hu, Jiaxu Miao, Jun Yu, Zhenyu Zhang, Shuohuan Wang, Yu Sun
Abstract: AI-generated image detectors deployed in open-world environments inevitably face distribution shifts as new and stronger generative models continue to emerge. Although existing methods improve cross-generator generalization through better representations or training data construction, they typically follow a static train-once-and-deploy paradigm and cannot adapt after deployment. In this work, we study open-set AIGC image detection from a test-time adaptation perspective. We propose Test-Time Curriculum (TTC), a simple and model-agnostic framework that adapts a detector on unlabeled test data through curriculum-based self-training. TTC starts from highly reliable pseudo-labeled samples and progressively incorporates harder yet informative cases, while enforcing class-balanced selection to reduce biased updates under generator shift. To further improve pseudo-label quality, we introduce Cross-Scale Pseudo-Label Refinement, which aggregates complementary evidence across multiple resolutions for more reliable adaptation, and applies noisy-or fusion at inference to strengthen final predictions. In addition, we construct AIGCGuard, a new benchmark containing 3,100 representative real images and 124,000 generated images from 40 of the most advanced open-source and proprietary text-to-image models. Extensive experiments on five benchmarks show that TTC substantially improves overall detection performance under diverse unseen-generator shifts, establishing a practical and effective test-time adaptation framework for open-set generated image detection.
Authors: Jinhua Zhang, Yisong Lin, Wei Long, Shuhang Gu
Abstract: Autoregressive image generators are commonly pretrained with token-level cross-entropy under teacher forcing, yet evaluated by the distributional quality of decoded images. This creates an objective mismatch, because categorical errors have unequal image-level consequences, and a context mismatch, because inference conditions on model-generated histories. We introduce FD-loss post-training, which adapts a pretrained discrete generator using representation-space Fr\'echet distance as the sole objective. A dual-pass scheme first constructs detached rollout contexts through gradient-free generation under the model's native inference configuration, then performs differentiable replay with a probability-level straight-through estimator (STE) that preserves hard argmax decoding in the forward pass while propagating image-level gradients through temperature-scaled probabilities. Only the generator is updated, while the tokenizer and feature extractors remain frozen. Across eight completed configurations from four generator families on class-conditional ImageNet at $256\times256$, FD-loss post-training reduces FID and $\mathrm{FD}_{r6}$ by 41.4% and 52.0% on average. The strongest FID result improves from 2.42 to 1.43 without adding parameters or inference steps.
Authors: Ziang Wu, Peng Jin, Qishen Yin, Munan Ning, Hao Li, Peizhen Zhang, Li Yuan
Abstract: Vision-language MoE batches contain different numbers of image and text tokens. Image resolution, image count, tiling, and prompt length all change this token mix. We call the standard token-level Switch auxiliary loss Std-Aux. Std-Aux balances only the mixed load, so large image and text load errors can cancel at one mix. On our main model, the same trained router shows more than a fivefold change in load imbalance across image resolutions. We hold the image and text load profiles fixed and derive the exact load curve as the token mix varies. The image-text load gap controls sensitivity to the token mix. Physical preprocessing can also change the conditional profiles. The fixed-profile law excludes such changes. To design a remedy, we examine the router input structure. Image and text occupy distinct regions, while visual tokens group strongly by source image. The modality boundary motivates separate image and text terms. The image boundary motivates one equal-weight routing instance per image. ReBA, or Relax Within, Balance Across, implements both choices. Across four split backbones, ReBA lowers load on every reported benchmark input while keeping mean task accuracy comparable to Std-Aux. ReBA also lowers average load over the tested range and worst physical load under resolution and tiling shifts. Code is available at https://github.com/ZiangWu-77/ReBA.
Authors: Jingtong Chen, Jiahui Wang, Xue Zhao, ShaoGuo Liu, Minghao Li
Abstract: Recent advances in image generation and editing have made prompt quality a key bottleneck for e-commerce creatives. Vision-language models (VLMs) can generate image-editing prompts from product images and metadata, but further improving their prompt-writing capabilities requires post-training with feedback from the generated images. Group Relative Policy Optimization (GRPO) is a natural framework for such outcome-level reward optimization. However, it assigns credit only at the full-prompt level, even though image quality often depends on specific design elements such as composition, background, and the presentation of selling points. Existing fine-grained credit assignment methods typically require step-level supervision or learned critics. To address this, we propose EAGLE-GRPO (Element-Aware Group Learning for E-Commerce Image Generation), which decomposes the group-centered reward over predefined elements. We cast element-level credit assignment as a kernel ridge regression problem and derive a closed-form solution, without additional rollouts or separate credit-assignment models. This yields interpretable per-element advantages and more precise policy updates. Experiments show that EAGLE-GRPO sustains performance gains over more training steps before plateauing and generates prompts that produce higher-quality e-commerce images than competitive VLM prompt-writing baselines.
Authors: Mingya Alexa Gong, Da Ma, Lovre Antonio Budimir, Ivana Matovinovic, Sven Loncaric, Myeong Jin Ju, Yukun Zhou, Siegfried K. Wagner, Pearse A. Keane, Marinko V. Sarunic
Abstract: Despite the widespread adoption of foundation models as feature extractors for medical imaging, relatively little is understood about how different pretraining strategies influence the transferability of learned representations to weakly supervised ophthalmic imaging tasks. We investigate this question in ultra-widefield (UWF) retinal imaging by evaluating foundation model representations within a patch-based multiple instance learning (MIL) framework for disease classification on UWF images. We compare Vision Transformer encoders pretrained with supervised, Masked Autoencoder (MAE), and self-distillation objectives, while keeping the downstream aggregation architecture unchanged. Within a controlled comparison of ViT-B encoders pretrained on ImageNet-1k, the choice of pretraining objective substantially influenced frozen representation transfer, with supervised and self-distillation-based models outperforming MAE. A contemporary DINOv3 model pretrained at a larger scale achieved the strongest overall performance, with a quadratic weighted kappa of 0.863 for five-class diabetic retinopathy grading, comparable with DINOv1. Attention analysis further revealed distinct patch-aggregation behaviours associated with the different pretrained representations, while partial fine-tuning substantially reduced the performance gap for MAE. These findings suggest that pretraining strategy influences both representation transferability and the subsequent aggregation of patch-level evidence within MIL, resulting in differences in downstream classification performance.
Authors: Chaoyue Wu, Yunfei Zhang, Si Wu
Abstract: Layout-to-image diffusion models have achieved impressive semantic controllability by conditioning generation on category-level segmentation maps. However, such category-aligned control is not necessarily instance-addressable: multiple nearby objects from the same category are often treated as a shared semantic region, leading to ambiguous boundaries, averaged appearances, and feature confusion among instances. This limitation is particularly evident in urban scene synthesis, where small and crowded pedestrians or vehicles require fine-grained instance separation while preserving global scene consistency. In this paper, we propose InstancePin, an instance-addressable layout-to-image diffusion framework that pins each object instance with an explicit coordinate anchor. Instead of directly injecting instance masks into the pretrained backbone, InstancePin introduces an independent instance-aware adapter to preserve the category-level generation prior while learning instance-specific spatial control. For each instance, its center coordinate is encoded with Fourier features and projected into a coordinate token, which serves as a spatial anchor queried by latent image features through coordinate pinning attention. To make these anchors spatially meaningful, we further supervise the coordinate attention maps with instance regions, encouraging each coordinate token to activate its corresponding object area. Finally, an instance-mask guided fusion module routes pretrained backbone features to non-instance regions and adapter features to instance regions, enabling local instance refinement without sacrificing global semantic fidelity. Extensive experiments on Cityscapes demonstrate that InstancePin mitigates instance entanglement in dense layouts and improves both image fidelity and semantic consistency.
Authors: Jian Xu, Yanning Wu, Delu Zeng, John Paisley, Qibin Zhao
Abstract: Many physical attributes are \emph{irreversible}: ice melts but does not re-freeze, paper chars but does not un-burn. Do video generators respect this? We show the question is hard to measure, and that what can be measured reliably is \emph{development} rather than reversal. Metrics of local reversal are null-degenerate: a per-clip violation rate scores $0.50$ on pure noise, and a variance-normalized reversal residual sits at its noise ceiling. What survives null-testing is a two-part protocol: progress (a directional attribute correlation) and a stasis rate. Under this protocol, generated video separates cleanly from real footage, and the gap is human-validated. Across seven text-to-video models, real reference footage advances ($\rho{=}{+}0.40$, $35\%$ static) while every generator shows near-zero progress and $92$--$100\%$ stasis; nine annotators rate real footage far above generated ($2.75$ vs.\ $0.99$ on a $0$--$4$ scale). The reliable finding is \emph{under-development}: generators barely advance irreversible attributes rather than reversing them. As a complementary mechanism, we show that post-hoc readout guidance is gameable, whereas enforcing monotonicity by construction in a disentangled attribute latent removes the gameable readout, validated in controlled and semi-synthetic settings.
Authors: Marcel Plocher, Bernhard Sch\"olkopf, Andreas Geiger, Gege Gao
Abstract: The target representation defines the distribution an image generator must learn, yet it is often treated as an interchangeable interface. This assumption is particularly questionable for continuous masked generators, which combine contextual inference from visible tokens with conditional modeling of each missing token. We study raw pixels, SD-VAE latents and DINOv2 as well as MAE representation-autoencoder features within a unified masked autoregressive rectified-flow model. Under a shared ImageNet training budget, these spaces exhibit distinct optimization and inference regimes. DINOv2 converges fastest in both iterations and computation but benefits strongly from a wider local denoiser and direct context fusion. Pixels optimize substantially more slowly and require a different prediction, masking, and guidance configuration. MAE reconstructs images more faithfully and exhibits clear semantic clustering, yet produces generations substantially worse than DINOv2. The representations also respond differently to classifier-free guidance and occupy distinct precision-recall trade-offs. Together, our results show that compression, reconstruction fidelity, token dimensionality, and visible semantic clustering do not individually predict generative behavior. Instead, target representations redistribute difficulty across contextual modeling, per-token denoising, and inference-time distributional control.
Authors: Alison M. Fernandes, Hermes I. Del Monego, Bruno S. Chang, Anelise Munaretto, H\'elder M. Fontes, Rui L. Campos
Abstract: Recently, Wi-Fi sensing has played a significant role in Human Activity Recognition (HAR), as it enables the detection of various activities using only Wi-Fi signals, ensuring privacy and remaining non-intrusive for the user. However, environmental characteristics such as reflective surfaces, hardware offsets, and other physical impairments affect recognition by the neural network, subsequently causing errors and significantly reducing model accuracy. To overcome this problem we present the WiFuse framework, a dual-stream Channel State Information (CSI) framework for human activity recognition (HAR) that pairs denoised time-domain amplitude variations with 2D-FFT-derived Delay-Doppler motion representations computed from the sanitized channel phase. The fused representation feeds a hybrid ResNet-Temporal Convolutional Network (TCN) neural architecture augmented with channel and spatio-temporal attention, where the ResNet extracts spatial-spectral features and the TCN models long-range temporal dependencies; a decoupled two-stage transfer learning strategy is employed to improve optimization stability and feature reuse. We conduct extensive experiments on two public datasets, including comparisons against state-of-the-art methods and alternative hybrid architectures, ablation studies, and cross-dataset and domain-adaptation evaluations. The proposed framework reaches an overall accuracy of up to 95.28% across the four environments of the XRF55 dataset and up to 98.20% on the multi-user Wi-MIR dataset. Overall, the results indicate that combining amplitude and Delay-Doppler representations within a dual-stream strategy, enhanced by transfer learning, improves recognition performance under conditions that typically degrade deep neural networks, such as class overlap, multipath propagation, noise, and interference.
Authors: Zhiheng Li, Lei Chen, Jie Zhou, Jiwen Lu
Abstract: Screen content images are generally composed of texts and graphics. Compared to natural images, these man-made images contain a large quantity of sharp but repetitive structures. However, existing works in screen content super-resolution underutilize the special characteristics of screen content, leaving a large room to improve model performance and speed up. In this paper, we propose PixelSR, a simple yet effective method to improve super-resolution performance but with faster inference speed. To improve model performance, we classify pixels via pixel binning to compute content attention in the training phase. Specifically, after binning pixels into content-dependent groups, content attention is aggregated from pixel features within each group to introduce a content-dependent and non-local receptive field for every pixel. In the testing phase, we utilize the properties of self-repetitiveness and redundancy in screen content to speed up inference without the loss of model performance. We divide targeted high-resolution pixels into three types, which are unique pixels, repeated pixels, and background pixels for each test image. We conduct conventional network processing on unique pixels and cache their predictions in the on-the-fly lookup table. For repeated pixels which have appeared in unique pixels, we directly retrieve prediction results from the lookup table without network processing. For background pixels, we use the nearest neighbor algorithm to generate high-resolution pixels. The on-the-fly lookup table is cleaned and repeats the procedure above for the next test image. Experiments show our PixelSR achieves state-of-the-art performance with shorter inference time in screen content super-resolution.
Authors: Chenggong Hu, Shaoyin Ma, Yi Wang, Li Sun, Mingli Song, Jie Song
Abstract: Audio-driven emotional talking face generation aims to synthesize realistic videos with expressive facial dynamics. However, existing methods struggle to balance controllability and visual fidelity. Although implicit representations capture rich semantics, they lack structural guidance, often resulting in averaged emotional expressions. In contrast, explicit geometric methods offer better control over facial expressions but tend to sacrifice high-frequency texture details. To address it, we propose GemTalk, a diffusion-based framework that combines the semantic richness of implicit representations with the structural precision of explicit geometric priors. We introduce a Vision-guided Audio Emotion Projection (V-AEP) module to extract implicit emotional lip and expression features. At the same time, a Diffusion-based Geometric Priors Generator (D-GPG) generates identity-aware blendshape coefficients as explicit structural priors. Crucially, our Geometry-guided Emotion Modulation (GEM) module leverages these geometric priors to recalibrate the magnitude of implicit features, enabling precise, continuous control over emotional expressions, especially emotion intensity, without sacrificing visual quality. Extensive experiments show GemTalk achieves superior performance in photo-realism, and facial emotional dynamics.
Authors: Peng Zheng, Ruiqi Liu, Rui Ma, Zuxuan Wu
Abstract: Recent subject-to-image models have achieved impressive progress in personalized image generation, yet they still struggle to preserve fine-grained subject-specific details. A major reason is the lack of high-quality fine-grained identity supervision: real paired data are expensive to collect, while synthesized training pairs often preserve only coarse subject appearance and fail to capture subtle subject-specific details. In this work, we propose CopyCat, a lightweight model-refinement framework that improves fine-grained subject consistency within only a few seconds. CopyCat performs a one-time refinement of a pretrained subject-to-image model by attaching a lightweight Fine-grained Consistency LoRA (FCLoRA) and optimizing it using a single proxy image, which is used as both the conditioning image and the reconstruction target. This exact self-reconstruction objective substantially simplifies the optimization task, enabling effective fine-grained refinement within only a few seconds. The refinement is performed only once; the resulting model can be directly applied to diverse unseen reference subjects and prompts without further subject-specific optimization. We further revisit subject-to-image LoRA training in double-stream diffusion transformers and find that adapting only the visual stream consistently improves subject consistency. Extensive experiments on DreamBench and XVerseBench demonstrate consistent improvements in fine-grained subject consistency across representative subject-to-image models under both single- and multi-subject settings.
Authors: Kaihua Tang, Ziqing Xia, Xiaoxu Zheng, Xiaoxue Zhang, Michael Bi Mi, Zhan Xu, Dave Zhenyu Chen
Abstract: Despite recent advances in Monocular Depth Estimation, state-of-the-art depth foundation models remain vulnerable to robustness issues. Particularly, even slight camera rolls can result in substantial degradation in depth estimations. We attribute this problem to a previously overlooked phenomenon, termed the Horizontal Prior, which is a manifestation of long-tailed distribution bias: most training images are captured in approximately horizontal orientations due to human visual preferences and photographic habits. While intuitive remedies such as re-balanced data augmentation and horizon leveling provide partial improvements, they fail to fully address the issue. In this paper, we introduce Invariant Depth Constraint (ID-Constraint), a training-time supervision strategy that improves roll robustness by fine-tuning and jointly regularizing the depth backbone with a series of geometric and spatial reasoning tasks. These auxiliary objectives encourage the backbone to learn rotation-stable, depth-relevant representations, while the auxiliary prediction heads are discarded after training, leaving the original inference architecture unchanged. Extensive experiments on five benchmark datasets across four roll settings demonstrate the effectiveness of the proposed method.
Authors: Wei Lu, Yi Liu, Si-Bao
Abstract: Shadows in remote sensing images obscure surface appearance and disrupt radiometric continuity, reducing the reliability of visual interpretation and downstream analysis. Remote sensing image deshadowing is an ill-posed inverse problem that requires spatially varying illumination recovery while preserving chromatic and radiometric consistency in non-shadow regions. Existing methods commonly rely on hard shadow masks for compensation or directly regress RGB intensities. Hard masks may inadequately model gradual penumbra variations and are sensitive to localization errors, often producing residual shadows or halo artifacts; direct RGB regression entangles illumination recovery with chromatic reconstruction and can introduce color casts. To this end, we propose the Boundary-Reliable Illumination-Color Interaction Network (BRIC-Net), which decouples these failures at different representation levels. A Lightness Reliability Prior (LRP) derives reliability-aware guidance from CIELAB statistics. Boundary-Adaptive Gated Mixing (BAGM) performs gated interpolation between shallow RGB and lightness features around uncertain transitions, while Spatial-Channel Mutual Modulation (SCMM) coordinates deeper spatial and channel responses for appearance-preserving illumination recovery. BRIC-Net achieves 29.46~dB full-image peak signal-to-noise ratio (PSNR) on AeroDS-Syn and 27.96~dB on SRGTA. It also obtains the lowest Perception-based Image Quality Evaluator (PIQE) scores on AISD and AeroDS-Real. Region-wise evaluations and component ablations further support its effectiveness in shadow recovery and non-shadow preservation.
Authors: Yihan Qiu, Xiaodong Lin, Baoquan Zhao, Hailong Jiao, Ge Li
Abstract: LiDAR point clouds provide explicit, deterministic physical boundaries critical for collaborative safety-critical perception. However, wireless channels inherently impair and corrupt transmitted signals. Existing robust frameworks (such as deep JSCC or MDC) attempt to counter these channel impairments through statistical or parametric estimation, turning exact physical measurements into unverified algorithmic estimates. To address this, we propose Proteus, a learned LiDAR codec operating on 2D range images. By decoupling the frame representation into independent coders for the \textbf{sig}nificant range bit-planes (SIG) and the \textbf{ins}ignificant range bit-planes and attributes (INS), Proteus achieves overall stream-level truncation robustness. The non-truncatable SIG block encodes the most significant range bit-planes to establish a necessary, self-contained perceptual lower bound, below which the reconstructed point cloud is severely degraded. Meanwhile, INS employs bit-plane slicing representation and coding, ensuring that range truncation mathematically maps to a deterministic spatial precision degradation. Subordinate attributes are reconstructed via a hybrid lossless-predictive method, leveraging the decoded geometry as a strong structural prior for fine-grained approximation. Furthermore, strategic ordering within INS prioritizes geometry over attributes under bandwidth drops. Experimental results on the Waymo Open Dataset and SemanticKITTI demonstrate that Proteus tolerates up to approximately 70\% bitstream truncation, while outperforming established standards (G-PCC, Draco, and JPEG XL) and the representative learned compressor Unicorn under ideal channel conditions.
Authors: Zhenyang Li, Zongqi He, Jia Pan, Shijie Lin, Yifan Peng
Abstract: Event cameras offer microsecond-level temporal resolution and high dynamic range, potentially facilitating motion-blur-free panoramic imaging from fast rotational scanning. Nonetheless, existing optimization-based methods remain computationally demanding, while prior learning-based reconstruction methods are largely designed for perspective imagery and lack geometry-aware support for panoramic outputs. We present E2Pano, a geometry-guided event-to-panorama pipeline with an end-to-end learnable photometric reconstruction stage. Our framework preserves real spherical coordinates from geometric mapping throughout the pipeline, employs a lightweight enhancement module with frequency-domain supervision to bridge the event-image domain gap, and leverages a spherical Transformer with 3D positional embeddings for photometric reconstruction. Experiments on synthetic data and captured rotational scans show improved reconstruction quality and lower photometric reconstruction cost than optimization-based baselines, together with encouraging transfer to real captures under our acquisition protocol despite training purely on synthetic data. Additionally, we construct PanoScan, a dataset with 4,370 synthetic and 30 real-world panoramic scenes paired with event streams. Our dataset and code will be released.
Authors: Jianfeng Qiu, Peiyuan Li, Juan Xie, Xueliang Ma, Sihang Zhou, Yanning Hou, Ke Xu
Abstract: Zero-shot anomaly detection (ZSAD) aims to detect anomalies and localize defective regions in unseen target domains without target training data. Recent ZSAD methods build on pretrained vision models, particularly CLIP, and construct normal and anomaly references from textual prompts or learnable visual representations. These methods perform anomaly discrimination primarily in spatial feature spaces, where subtle changes in texture, boundaries, and local structures can be confused with normal appearance variations. Although inconspicuous spatially, such defects can disrupt local texture regularity or boundary continuity, inducing response deviations across frequency bands. However, existing ZSAD methods do not explicitly model these frequency-dependent characteristics. Our image-domain analysis reveals that local defects exhibit spatial-frequency deviations from normal references across low-, middle-, and high-frequency bands, indicating that anomaly evidence is not universally dominated by high-frequency responses. Motivated by this observation, we propose FreqAnchorAD, a frequency-aware framework that organizes frequency-enhanced responses for anchor-relative anomaly discrimination. Specifically, the Local Frequency Compensation Module (LFCM) enhances intermediate patch tokens with local spatial-frequency cues. The Frequency-Deviation Anchor Projector (FDAP), our core discrimination module, organizes enhanced responses along a source-derived channel coordinate and measures anomaly evidence through relative similarity to normal and anomaly anchors. Finally, Asymmetric Anchor Supervision (AAS) stabilizes normal-anchor alignment while preserving diverse anomaly patterns. Experiments on thirteen industrial and medical benchmarks show that FreqAnchorAD achieves state-of-the-art mean performance in image-level anomaly recognition and pixel-level defect localization.
Authors: Wei Lu, Hongyuan Liu, Si-Bao Chen
Abstract: Nighttime aerial image enhancement is challenged by spatially nonuniform exposure, mixed illumination, and weak structural evidence, while registered normal-light targets are difficult to capture from moving platforms. Generated normal-light images provide practical appearance guidance but may alter geometry or texture. We introduce \aeronight{}, comprising 1,500 real nighttime aerial RGB images: 1,300 inputs are associated with manually screened pseudo-references, and 200 inputs support unpaired evaluation. We propose AeroLLE, a two-stage framework that first recovers visibility with an HVI Base Enhancer and then performs Spatially Adaptive Exposure--Color Calibration (SAECC). After the Base Enhancer is selected and frozen, SAECC predicts bounded, low-resolution RGB gain and bias fields, restricting the magnitude and spatial variation of the second-stage correction. Experiments under complementary pseudo-paired and unpaired protocols demonstrate improved agreement with screened appearance targets, together with more balanced exposure and color correction across diverse nighttime aerial scenes. These results support constrained, stage-specific calibration as a practical strategy for learning from generated appearance guidance when registered aerial references are unavailable.
Authors: Junyang Zhang, Puhan Luo, Chen Tang, Yuxi Shi, Xiang-Yang Li
Abstract: Recent advances in large vision-language models (LVLMs) have enabled long-video understanding and analysis. However, processing the large number of frames in a video incurs substantial computational overhead. Existing methods reduce LVLM inference costs by scoring frame-query relevance before inference and selecting keyframes accordingly. Nevertheless, the distribution of relevant frames varies across queries, and these methods often need to score hundreds or thousands of frames. To address this limitation, we propose CSES, a training-free semantic keyframe selector that adaptively determines the numbers of frames to score and keyframes to select. CSES estimates the prominence of the frame-query relevance profile to guide active acquisition and adapt the temporal coverage of each input. It then formulates keyframe selection as a coverage problem that jointly accounts for semantic relevance, temporal redundancy, and visual redundancy. Active acquisition and keyframe selection terminate based on coverage saturation. The selection objective is monotone and submodular, enabling greedy optimization with a standard approximation guarantee. Experiments with four LVLMs on two benchmarks show that our method preserves accuracy while scoring $4$-$13\times$ fewer frames and selecting $18.4\%$-$20.5\%$ fewer input keyframes than existing baselines. CSES further achieves a $3.1$-$5.4\times$ speedup in frame selection over baselines.
Authors: Jun Nie, Yonggang Zhang, Tongliang Liu, Yiu-ming Cheung, Bo Han, Xinmei Tian
Abstract: Robust detection of generated images is critical to counter the misuse of generative models. Existing methods primarily depend on learning from human-annotated training datasets, limiting their generalization to unseen distributions. In contrast, large-scale vision models (LVMs) pre-trained on web-scale datasets exhibit exceptional generalization power through exposure to diverse distributions, offering a transformative paradigm for this task. However, our experimental results reveal that LVMs pre-trained on natural-image-dominated data can effectively capture the features of both natural and generated images, yielding comparably low losses and thus limited discriminative capacity between them. This prompts a key question: When and how do LVMs exhibit different behaviors when capturing features of natural and generated images? This investigation reveals an insight: during unlearning, LVMs exhibit disparate forgetting dynamics with feature degradation for generated images escalating faster than natural ones. Inspired by the disparate dynamics, we introduce two detection methods: 1) data-free detection, which prunes model parameters to induce unlearning without data access, and 2) data-driven detection, which optimizes LVMs to unlearn knowledge tied to generated images. Extensive experiments conducted on various benchmarks demonstrate that our unlearning-based approach outperforms conventional detection methods. By recasting the detection task as a problem of machine unlearning, our work establishes a new paradigm for generated image detection.
Authors: Mateusz Michalkiewicz, Mahsa Baktashmotlagh, Guha Balakrishnan
Abstract: Frozen vision foundation models are commonly evaluated through a single global image embedding, but this interface can conflate missing information with information lost at readout time. We study this distinction by keeping a pretrained vision encoder frozen and varying only the readout applied to its final patch tokens. We compare standard global readouts against a lightweight foveated readout, which attention-pools patch tokens using a learned or question-conditioned query, and against an oracle readout with access to the annotated target region. We evaluate these interfaces on three localized binding problems: a controlled synthetic color--shape binding task under clutter, a color-free crowded shape-detection variant, and a GQA-derived natural-image task where paired questions ask for the colors of different same-category objects in the same image. Global readouts perform near perfectly when the synthetic target appears alone, but collapse under clutter and counterfactual target edits, whereas the foveated readout recovers most of the oracle-accessible signal. On the GQA-derived task, question-independent global image vectors improve only modestly over question-only priors, while question-conditioned foveation substantially improves paired localized color accuracy. A counterfactual nuisance-to-signal ratio explains the synthetic failures: global pooling dilutes localized label-changing evidence while exposing the probe to nuisance variation from irrelevant objects. These results indicate that apparent spatial blindness in frozen vision models can arise from the global embedding interface rather than from an absence of spatial information in the frozen patch tokens.
Authors: Mridula Vijendran, Shuang Chen, Hubert P. H. Shum
Abstract: Restoring severely degraded visual media still remains a formidable challenge, as existing methods often hallucinate unnatural textures and contents, struggle with preserving color and texture, or fail to leverage partially retained image information. Existing restoration benchmarks assume known degradation operators and fail to capture the complex characteristics of artistic damage such as cracks, stains, and color/texture deviation. We introduce a controlled benchmark for blind restoration of semantic, semi-transparent image media degradations, accompanied by a new, publicly open degradation alpha texture mask dataset MDTD-Art. We present a new dataset and benchmark evaluating state-of-the-art universal restoration models against image editing and vision-language models across varying mask opacity levels. Our experiments demonstrate that image editing models consistently outperform specialized restoration architectures for arbitrary degradations, with performance gains amplified by structured prompt engineering emphasizing detail preservation and structural consistency. These findings position recoverable semantic information and prompt controllability as critical factors in art image restoration.
Authors: Jiaxuan Kang, Siyu Chen, Mingda Li, Mingjie Liu, Tianyue Wang, Zhaoyang Wei, Yongheng Zhang, Yanchao Hao, Zheng Wei
Abstract: Multimodal large language models have advanced visual understanding, yet perception-intensive reasoning remains challenging. Recent latent visual reasoning methods introduce hidden-space computation before answering, but they often rely on costly intermediate supervision, such as bounding boxes, sketches, or interleaved rationales. These strategies focus on how latent states should be shaped, but do not explicitly assess whether the latent is useful for the final answer. We propose LUT, a latent reasoning framework trained with only standard VQA pairs. LUT centers training on Latent Utility at two levels. At the trajectory level, we propose Utility-Aware Latent Distillation SFT, which explores answer-relevant latent trajectories, selects qualified trajectories by their information gain, and distills more reliable and learnable supervision through curriculum learning. At the step level, we propose Latent Attribution Policy Optimization, which uses answer-to-latent attribution to differentially optimize latent steps during reinforcement learning. Experiments on perception-intensive visual reasoning benchmarks show that LUT outperforms previous latent reasoning methods and remains competitive with latent-text interleaved methods with lower annotation cost.
Authors: Nil Stolt-Ans\'o, Maik Dannecker, Steven Jia, Julian McGinnis, Daniel Rueckert
Abstract: Clinical acquisition in cardiac magnetic resonance (CMR) imaging involves obtaining cross-sectional planes of the heart along the radial and longitudinal directions. Despite these planes being 2D cross-sectional images of the heart, radiologists understand the 3D spatial and continuous temporal nature of the organ being imaged. The same can not be said about the conventional deep learning architectures used to process CMR images, which rely on in-plane and grid-based operations, and are hence unable to organically integrate information from all imaging planes. This paper builds upon previous work on neural implicit segmentation functions (NISF) to overcome unaddressed challenges in cardiac function modeling in the CMR domain. For a given subject, our architecture builds a shared 3D+time representations from all available acquisition planes regardless of orientation. By design, predictions along any imaging plane orientation are cross-sections of the same 3D representation, leading to spatio-temporal consistency across all slices. Moreover, our architecture makes the rotation and translation parameters of imaging planes learnable, allowing us to correct for the commonplace respiratory and patient motion between slice acquisitions under a rigid assumption. Furthermore, interpolation of intensities and segmentation can be performed in 4D at any desired resolution. We perform our study on a 120 subject sub-cohort of CMR imaging data from the UK-Biobank. Our in-plane segmentation performance is on-par with existing CMR segmentation methods and explore how the majority of failure cases arise from limitations in the ground-truth segmentation, for which our representations make predictions with better anatomical accuracy than its original training data. We also evaluate our motion-correction capabilities, displaying quantitative and qualitative improvements in slice alignment.
Authors: Zhiqiang Lao
Abstract: One-step text-to-image models enable training-free, inversion-free editing with only 1--2 network function evaluations (NFE), while ChordEdit stabilizes such edits through low-energy smoothing along sampling time. Applied independently to video frames, however, it produces temporal flicker and edit-strength drift. We introduce \textbf{ChordVideo}, which extends the same low-energy principle to video time through shared noise, motion-aligned causal aggregation of per-frame Chord fields, and an optional temporally smoothed proximal correction. We derive a warping-error bound that separates motion bias from stochastic flicker and predicts diminishing returns with larger temporal windows. On TGVE/DAVIS with two one-step backbones, ChordVideo reduces warping error by \textbf{78\%} and flicker by \textbf{49\%}, improves CLIP frame consistency by \textbf{9--10 points}, and increases background PSNR by about \textbf{1.5,dB}, while retaining \textbf{2 NFE/frame}. Compared with seven multi-step editors, it achieves competitive temporal consistency and source preservation using \textbf{10--60$\times$ fewer model steps per clip
Authors: Soslan Kabisov, Gennadiy Savrasov, Maksim Elistratov, Antonio Rodriguez, Daniil Ignatiev, Nikita Gavrilov, Rustam Uzdenov, Alexey I. Boyko, Igor Pasechnik, Anton Konushin, Andrey Kuznetsov, Dmitrii Zhemchuzhnikov
Abstract: Computer-Aided Design (CAD) underpins modern engineering, yet converting existing shapes into editable models still demands substantial expert effort. Most AI systems emit the entire CAD program in a single pass, never inspecting the intermediate geometry. In contrast, human engineers build a part feature by feature, checking after each operation what remains to be modeled. We introduce CADENA (Spanish for "chain"), a model that reconstructs a 3D mesh as a parametric CAD program, growing its sequence of operations one at a time and comparing the target with the currently predicted geometry at every step. We also address the lack of benchmarks for evaluating reverse-engineering methods on mechanical parts, introducing CADENA-Bench, a benchmark that measures performance across categories of mechanical parts. CADENA outperforms prior methods on CADENA-Bench and on the DeepCAD, Fusion 360, and MCB datasets. Code is available at https://github.com/zhemdi/cadena, model weights at https://huggingface.co/kulibinai/cadena, and CADENA-Bench at https://huggingface.co/datasets/kulibinai/cadena-bench.
URLs: https://github.com/zhemdi/cadena,, https://huggingface.co/kulibinai/cadena,, https://huggingface.co/datasets/kulibinai/cadena-bench.
Authors: Pengyun Qiu, Shuo Wang, Zeyuan Chen, Yihao Zhi, Chongjie Ye, Xiaoguang Han
Abstract: Injection molding is the cornerstone of mass-producing plastic components. While current algorithms can automate mold design for basic geometries using standard two-piece molds, complex parts featuring undercuts, side holes, or re-entrant features present a significant challenge. These geometries often necessitate auxiliary components beyond the primary upper and lower molds. In practice, designing these intricate assemblies is a laborious process that relies heavily on expert knowledge. Furthermore, the scarcity of public datasets has hindered the development of effective learning-based solutions. To bridge these gaps, we introduce MoldCAD, a curated dataset that pairs complex single-body CAD parts with industry-standard mold assemblies. Each entry includes the upper and lower molds, parting surfaces, demolding orientations, and necessary auxiliary components. The dataset comprises 4,934 CAD models and over 3,850 mold assemblies, totaling more than 23k individual models. Building upon this dataset, we propose a comprehensive pipeline that predicts demolding orientations, identifies auxiliary components, and constructs parting surfaces to derive a complete, manufacturing-ready mold assembly for downstream CAD/CAM workflows. Our results demonstrate a promising path toward fully automated industrial mold design and contribute to the broader advancement of manufacturing-aware CAD generation.
Authors: Wenrui Cai, Yuzhe Li, Qingjie Liu, Yunhong Wang
Abstract: Most current visual trackers adopt a matching-based architecture trained exclusively on tracking datasets, whose performance gains depend heavily on the length of the input context, and have now reached a bottleneck. While high-performance tracking increasingly relies on foundation models, existing methods use them monolithically, adapting a foundation model into a tracker or modify a segmentation foundation model into a tracking pipeline, which fails to exploit complementary strengths. Matching-based trackers excel at instance-level correspondence but lack semantic discrimination and fine-grained foreground perception, whereas segmentation foundation models produce precise masks yet struggle with instance discrimination and multimodal extension. Both paradigms also lack error-correction capabilities for long-term tracking. To address these issues, we propose ACTrack, an agentic coordination framework that treats heterogeneous models as invocable tools under an event-triggered mechanism. ACTrack coordinates a Tracker-based Instance Matching Tool for target discrimination, a SAM3 Motion Tool for mask-derived motion priors, a SAM3 Perception Tool for detecting distractors and instance-conflict cues, and a VLM Reprompt Tool activated only under persistent conflict to mitigate error accumulation. We design a complete tool-invocation trigger mechanism and an inter-tool coordination mechanism, enabling the complementary strengths of different model tools to be fully integrated. Experiments show that ACTrack substantially surpasses the strongest and the largest trackers on eight RGB benchmarks. Furthermore, a parameter-efficient adaptation strategy enables parameter sharing and reuse across tools, achieving unified multimodal tracking with only 30\% trainable parameters while substantially outperforming prior methods on multimodal benchmarks such as LasHeR, VisEvent, TNL2K, and DepthTrack.
Authors: Rebeka Popek, Vaghawan Ojha, Young Hwan You
Abstract: Many open educational resources are lacking in accessibility, especially in-depth image descriptions. In subjects like Science and Mathematics, however, it can be particularly difficult to write image descriptions since there can be many complicated expressions and names depending upon the course level. To help fill that gap in a small way, we introduce Math Image Descriptions for Accessible Learning (MIDAL), a math image-description dataset of 2,020 mathematical images spanning multiple educational levels, to aid in training vision language models to create image descriptions following accessibility best practices. We hope MIDAL is a valuable resource in enhancing the conversation and innovation regarding accessibility of STEM content in higher education. This dataset is however not just limited in math description generation but can also be used to fine-tune language models that can have improved mathematical reasoning and answers.
Authors: Xuechen Li
Abstract: Panoptic crop mapping requires both delineating individual agricultural parcels and assigning a crop type to each parcel from satellite image time series. Existing approaches typically rely on dense parcel-level annotations and task-specific model training, which limits their applicability to new regions and growing seasons. We introduce PhenoStitch, a panoptic crop-mapping pipeline that requires no task-specific gradient-based training. A frozen Segment Anything model first oversegments each patch into class-agnostic regions. For each region, optical NDVI and Sentinel-1 backscatter series are summarized by an analytic double-harmonic phenological signature. Adjacent regions are then merged into parcels by minimizing a Potts graph energy, and each parcel is classified by nearest-prototype matching using only (k) labeled parcels per class. A final topology-closure step produces the panoptic map. Under a matched budget of (k=20) parcels per class, corresponding to less than 1% of the available labels, PhenoStitch achieves 20.0 crop mIoU, 76.2 segmentation quality, and 6.2 panoptic quality on PASTIS-R under a 5-fold, 3-seed evaluation. It outperforms the evaluated frozen foundation-model, few-shot, and matched-budget supervised baselines under the same protocol, with a consistent ranking also observed on ZueriCrop. Ablation studies show that radar observations contribute the largest performance gain, while the graph-energy merge and compact phenological signature provide further improvements. These results demonstrate the effectiveness of combining label-free parcel delineation with few-shot phenological recognition for panoptic crop mapping under limited supervision.
Authors: Kevin Bui, Adina Ciomaga
Abstract: Robust to intensity inhomogeneity, the local Chan--Vese (LCV) model extends the classical Chan--Vese (CV) image segmentation method by incorporating local statistical information around each pixel. Originally, the LCV model was solved using a finite difference scheme, following the approach used for the CV model. As an alternative to the finite difference scheme, a more efficient algorithm based on the Merriman-Bence-Osher (MBO) scheme was later developed for the CV model. In this paper, we derive a similar MBO-based algorithm to solve the LCV model and propose an efficient implementation. The algorithm is developed for both two-phase and multiphase segmentation, and an extension to color images is also discussed. To demonstrate the effectiveness of the proposed approach, we apply it to a variety of grayscale and color images, including medical and microscopy images.
Authors: Dongheng Lin, Jianbo Jiao
Abstract: In animation production, paint-bucket colourisation for hand-drawn animation is a labour-intensive procedure that assigns each enclosed region in line sketches a colour from reference design sheets. Recent automatic paint-bucket colourisation pipelines mirror this workflow via region correspondence, but correspondences can be brittle when regions are ambiguous fragments without proper context. In this paper, we propose Palette Context Assisted (PeCA), a new training-free, plug-and-play framework for animation video colourisation that aims to close this gap at test-time via reasoning over spatial and temporal contexts. Extensive experiments on existing benchmarks and a newly introduced long-video test case show consistent performance boosts.
Authors: Zhuang Xiong, Guohao Zhang, Chen Zhang, Zheyu Jiang, Yuchao Mei, Qingshan Xu, Wenbing Tao
Abstract: Monocular panoramic SLAM benefits from substantial visual overlap under large camera rotations, yet remains prone to errors caused by camera tilt, scale drift, and false loop closures. We show that a frozen panoramic geometry foundation model provides useful internal cues beyond its explicit geometric outputs: intermediate tokens encode gravity in the camera frame, while cross-view attention provides a compatibility cue for potential revisits. Building on these cues, we present HALO-SLAM. A gravity readout enables IMU-free spherical upright canonicalization. For loop closure, we introduce a cost-aware three-stage cascade combining DBoW2 event-level retrieval, attention-based compatibility filtering, and dense geometric validation through symmetric submap augmentation. Accepted revisits yield pixel-aligned 3D--3D correspondences in both local gauges, from which robust $\mathrm{Sim}(3)$ constraints are estimated and jointly optimized with sequential constraints in a global pose graph. Across 125 sequences from five real-world panoramic benchmarks, our method achieves \textbf{100\%} sequence success (\textbf{125/125}) under the stated criterion and the lowest ATE among the evaluated methods on all five benchmarks, reducing ATE by \textbf{30--88\%} relative to the best ERP-native baseline on each benchmark.
Authors: Minseong Kweon, Junaed Sattar
Abstract: We propose Swimm3R, a unified framework that combines medium-aware structure-from-motion (SfM) with Underwater Beta Splatting to address scattering- and attenuation-induced failures in underwater 3D reconstruction. Swimm3R distills in-air geometric priors into a feed-forward backbone and uses a physics head to regress underwater image-formation parameters, camera poses, and restored point clouds. Additionally, we introduce Underwater Beta Splatting, which extends Gaussian splatting with Beta primitives and scattering-aware geometric gradients for stable underwater geometry representation. We further establish the Barbados underwater video dataset to demonstrate the effectiveness of our method in challenging underwater environments. On this dataset, Swimm3R robustly recovers underwater scene structure under challenging scattering conditions, yielding coherent seafloor geometry. Using these predicted point clouds, the proposed Underwater Beta Splatting improves average PSNR by $1.47$ dB over WaterSplatting while increasing downstream localization performance by $2.0$ and $2.4$ percentage points in RRA@15 and RTA@15, respectively.
Authors: Jiang Wu, Sichao Wu, Yinsong Ma, Lifang Zheng, Jingliang Duan
Abstract: Large vision--language models (VLMs) can reason step by step about complex visual scenes, but this open-ended, autoregressive chain-of-thought (CoT) approach is poorly suited to safety-critical, rule-governed settings such as industrial surveillance, where decisions must be bounded, deterministic, and low-latency. Because CoT inference cost scales jointly with reasoning length and the number of concurrent streams, it creates a throughput bottleneck that precludes the real-time, multistream monitoring required for industrial accountability. Here we present MonitorVLM-v2, a deployment-oriented framework that recasts VLM-based safety assessment as probabilistic inference over a finite regulatory decision space, compressing multimodal reasoning into single-step rule-ID predictions and reducing decoding from a variable-length sequence to a single token. We introduce symbolic policy optimization (SymPO), a novel contrastive policy optimization algorithm that sharpens decision boundaries within this finite symbolic space, together with an entropy-driven triage mechanism that routes uncertain predictions to human reviewers for expert confirmation. In a four-month prospective deployment across 10 concurrent camera feeds in an operational underground mining facility, MonitorVLM-v2 achieved a 19.45-fold increase in inference speed and identified 2.78 times as many confirmed violations as the site's routine manual inspection workflow, demonstrating the practical value of compressed symbolic decision-making for real-time, auditable industrial monitoring.
Authors: Myeongkyun Kang, Yanting Yang, Xiaoxiao Li
Abstract: Fine-grained visual representations are essential for medical image analysis, particularly when diagnostically relevant evidence is subtle and spatially localized. Modern transformer-based medical vision encoders must therefore learn patch-level representations that are both clinically meaningful and spatially consistent. Without these properties, large vision-language models (LVLMs) operate on an ambiguous visual foundation, limiting their ability to generate clinically reliable and spatially grounded responses. However, existing training strategies for medical vision encoders rarely achieve both objectives. Image-text alignment provides clinically meaningful supervision primarily at the image level, leaving the spatial localization of diagnostic evidence weakly constrained. In contrast, self-supervised learning promotes spatial consistency but lacks the semantic supervision needed to distinguish visually similar yet clinically distinct regions. To address this gap, we present LoFi, a medical vision foundation model built on location-aware fine-grained representation learning. LoFi trains a vision encoder with a lightweight large language model under grounding and grounded captioning objectives. Because these objectives require predicting location from clinical text and vice versa, spatial consistency emerges without any explicit patch-level regularization. To enable training at scale, we construct MedG, a large-scale medical grounding dataset of 4.48M image-text-box triplets curated from 84 datasets spanning 7 modalities. Across phrase grounding, visual question answering, and region-based organ classification under perturbations, LoFi consistently outperforms general-purpose and medical vision foundation models as well as state-of-the-art LVLMs. Code is available at https://github.com/myeongkyunkang/lofi-medg.
Authors: Kaifang Long, Lianbo Ma, Liming Liu, Guoyang Xie
Abstract: Current advancements in Multimodal Anomaly Detection (MAD) are largely driven by enhancing multimodal fusion, particularly through the integration of RGB and Depth data for richer anomaly representation. However, less attention was devoted to analyzing the role of cross-modal fusion bias, a well-known challenge in multimodal learning, in MAD. This gap motivates a key question: can we overcome this bias to break the performance bottleneck of current work? In this paper, we first analyze the impact of cross-modal fusion bias in MAD via the Fisher Information Matrix. Then, grounded in these findings, we propose UCFB, a simple yet effective plug-and-play framework designed to mitigate cross-modal fusion bias in MAD. It achieves this by jointly employing Fisher-information-guided dynamic calibration to adjust modality-specific regularization weights and canonical similarity analysis to improve inter-modal interactions. Extensive experiments on the MVTec 3D-AD and Eyecandies datasets demonstrate that UCFB achieves consistent improvements in single-class, multi-class, and few-shot settings.
Authors: Zhiyue Liu, Wenkai Zhou, Jian Qin, Qipeng Jiang
Abstract: Zero-shot image captioning aims to generate image descriptions without annotated image-text pairs. Recent approaches exploit text-to-image models to synthesize training data from text-only corpora, but most focus on improving overall data quality. In contrast, we observe that synthetic image-text misalignment is often structured and fine-grained: pairs may remain globally plausible while containing missing entities or misgrounded attributes, thereby degrading supervision fidelity. As a result, methods based on global similarity for image rematching or regeneration may improve apparent plausibility, but cannot systematically repair entity-level misalignment. To address this issue, we propose ReCap, a plug-and-play framework that shifts synthetic data refinement from implicit global matching to explicit fine-grained realignment. Specifically, ReCap enforces entity-level correspondence by using detected image-supported entities to guide caption rewriting, yielding more faithful synthetic supervision. In addition, we introduce an adaptive dynamic weighted learning strategy to downweight unreliable synthetic pairs during training. As a general framework, ReCap can be integrated into existing synthetic-data pipelines. Extensive experiments show that ReCap consistently improves image-text consistency and achieves state-of-the-art performance on both in-domain and cross-domain zero-shot image captioning benchmarks.
Authors: Timothee Mickus, Claudio Savelli, Eduardo Cal\`o, Emilio Raimond, Stella Frank, Hengyu Luo, Flavio Giobergia, Vincent Segonne, Chuyuan Li, Aman Sinha, Lorenzo Vaiani, J\"org Tiedemann, Ra\'ul V\'azquez
Abstract: In an age of rapid model turnover, how do we make hallucination evaluation more perennial? We explore whether human-written hallucination samples could take the place of model-generated hallucinations, in order to make benchmarking detection independent of particular models. To this end, we construct a dataset of 1,600 human-written samples, spanning four languages (Chinese, English, French, Italian), and 18,400 samples from five vision-and-language models, all annotated for hallucinations using a fine-grained span-level labeling scheme. We find that human-written samples result in higher agreement and allow greater control of dataset contents, while remaining distributionally similar to samples derived from vision-and-language samples and providing a reasonable portrayal of detection capabilities - suggesting that human data is a viable substitute for model-based hallucination benchmarks.
Authors: Han Jiao, Jiakai Sun, Lei Zhao, Wei Xing, Huaizhong Lin, Zhanjie Zhang, Ao Ma
Abstract: Constructing photorealistic Free-Viewpoint Videos (FVVs) of dynamic scenes from a set of posed 2D images has been an intriguing yet challenging task in computer vision. Methods based on neural rendering achieve high-fidelity image quality in FVV construction. However, most of these methods are unable to achieve real-time rendering and often require complete video sequences to train. Despite the existence of some online training methods capable of rendering FVVs in real time, they struggle to meet the requirements for storage and training time for downstream applications. To overcome this problem, we propose Struct-GStream, which can achieve efficient FVV streaming using structured 3D Gaussians (3DGs). Specifically, we introduce dynamic anchor points to generate structured 3DGs to construct basic scenes and model approximate scene movements based on the assumption of local rigidity in object motion. Besides, we introduce a global free 3DGs patching strategy involving free 3DGs' generation, pruning, and optimization to patch and model deficient areas and emerging objects. Our method achieves fast training at low bitrates while maintaining high rendering quality. Extensive experiments demonstrate that Struct-GStream significantly outperforms existing online training methods for FVV construction in terms of training time, storage, and rendering quality while maintaining competitive rendering speed.
Authors: Xinheng Han, Jianfei Wang, Yu Chen, Xiang Wang, Shuai Li, Weixing Li, Feng Pan
Abstract: Multimodal Large Language Models (MLLMs) are increasingly expected to solve structured perception tasks that require visual recognition, language-to-object binding, object cardinality preservation, and precisely localized grounding and segmentation outputs. However, existing group-relative reinforcement learning methods provide only response-level supervision, creating a granularity mismatch for structured multi-object prediction: a single advantage is broadcast to all tokens in a response, without distinguishing individual box contributions. To address this mismatch, we propose MCR-GRPO, a marginal contribution assignment framework that derives box-level credit directly from each sampled response. Specifically, Marginal Contribution Reward (MCR) estimates each predicted box's contribution through a leave-one-out comparison, measuring how the matched set value changes when the box is removed from the response. After within-response normalization, records that improve the set value receive positive credit, while redundant or harmful ones are suppressed. To make marginal attribution stable and informative, we further introduce a Continuous Matched Set Value Evaluator that integrates permutation-invariant matching, count-aware normalization, and graded localization. MCR-GRPO maps normalized box-level marginal advantages to the token spans that generated each box, preserving GRPO's response-level comparison while enabling box-aware optimization of structured multi-object grounding. Experiments across REC, DOD, segmentation, and counting benchmarks show state-of-the-art performance over prior GRPO-based baselines.
Authors: Huu Phong Nguyen, Shekhar Madhav Khairnar, Ganesh Sankaranarayanan
Abstract: Artificial Intelligence is increasingly applied to surgical video analysis for phase segmentation, skill assessment, and workflow optimization. A key challenge is the length of surgical recordings, often one to several hours, creating substantial computational burden. We previously developed Kinematics-Adaptive Frame Recognition (KAFR) for robotic surgery, showing that tracking tool motion effectively identifies informative frames while filtering redundant content. However, laparoscopic surgery introduces additional challenges: manual camera control causes frequent motion artifacts, and image quality is generally lower than robotic systems. This study evaluates whether KAFR generalizes to laparoscopic surgery using the Cholec80 benchmark, comprising 80 laparoscopic cholecystectomy procedures annotated for seven surgical phases. KAFR operates in three stages: a fine-tuned YOLO model detects and segments surgical tools; frames are adaptively selected based on tool displacement or velocity variation; and an X3D model classifies selected frames into surgical phases. KAFR achieved a 91.0\% F1 score using only 0.58\% of frames for phase classification, representing an approximately seven-fold reduction compared to typical 4\% frame sampling, while maintaining performance comparable to LoViT (90.2\%) and Trans-SVNet (89.7\%). These results demonstrate that kinematics-based frame selection transfers effectively to the challenging laparoscopic environment.
Authors: Ruijie Fan, Junyan Ye, Qi Zhu, Weijia Li
Abstract: Cross-view geo-localization (CVGL) retrieves geo-tagged satellite imagery for a ground-view query. Most systems exhaustively search a flat, fixed-resolution gallery, incurring high cost over large areas and adapting poorly to satellite resolution changes. Autoregressive coarse-to-fine alternatives reduce comparisons but bind later predictions to earlier decisions and a predefined hierarchy. We introduce GeoMoE, a sparse mixture-of-experts dual encoder that decouples global multi-scale representation learning from local hierarchical search. Global multi-scale supervision and content-adaptive routing map ground and satellite images across resolutions into a globally comparable embedding space. At inference, each image is encoded once, and probabilistic beam search follows parent--child links to score a small candidate subset. Later levels reuse these descriptors rather than features generated by preceding levels, limiting feature-level error propagation and hierarchy coupling. We further introduce VIGOR-M, a four-city benchmark with an explicit parent--child satellite hierarchy and held-out half-step galleries for single-resolution, cross-resolution, and hierarchical evaluation. GeoMoE achieves 95.78% R@40m on Just Zoom In, 2.77 percentage points above the previous best, and 62.39% R@1 on VIGOR-M. The latter requires 0.885 MMAC/query for descriptor matching, 5.27% of an exhaustive L3 scan, while exceeding the strongest exhaustive baseline by 3.12 percentage points in R@1. One model trained on L1, L2, and L3 also outperforms a matched dense control across all six galleries and transfers to three withheld resolutions. By decoupling globally trained embeddings from local hierarchical search, GeoMoE jointly improves localization accuracy, search efficiency, and cross-resolution transfer.
Authors: Hashmat Shadab Malik, Toluwani Aremu, Samuele Poppi, Muzammal Naseer, Salman Khan
Abstract: Training-free test-time defenses offer a practical way to improve the adversarial robustness of CLIP-style vision--language models without modifying the pretrained model. However, their correction strength is typically fixed for a narrow range of attack budgets, even though the attack budget is unknown at inference and the required correction varies across samples. We show that this mismatch causes existing defenses to degrade sharply as attacks strengthen. We introduce ReACT-CLIP, a response-conditioned test-time defense that separately determines how strongly each input should be corrected and whether defensive intervention is necessary. Our key observation is that the relative increase in CLIP visual-feature drift between low- and high-noise probes provides a graded, sample-specific proxy for correction demand. ReACT-CLIP maps this relative cross-noise drift to the Gaussian noise scale used to construct a stable, noise-averaged feature anchor, enabling the corrective reach to adapt to each input. To determine whether intervention is necessary, we further observe that clean inputs retain stable class-probability distributions under weak spatial augmentations, whereas adversarial inputs exhibit greater variation. ReACT-CLIP quantifies this variation using a prediction-instability score computed by Jensen--Shannon divergence and combines it with relative cross-noise drift to form the defensive intervention score. ReACT-CLIP requires no model or prompt training, and its correction-strength mapping is calibrated once and fixed across datasets and attack budgets. Across 12 downstream datasets, as well as ImageNet and its distribution-shifted variants, ReACT-CLIP delivers substantial robustness gains across diverse attack types and strengths while largely preserving clean accuracy.
Authors: Nathan Hu, Yang Yang, Fumio Okura
Abstract: Autoregressive rigging models such as UniRig and SkinTokens perform well on articulated characters, but their ability to generalize to plant structures remains largely unexplored, since plant topologies exhibit highly variable, non-canonical branching patterns that challenge learned skeletal priors. We evaluate these models for plant skeletal reconstruction using synthetic L-system-generated trees and real scanned data spanning monopodial, sympodial, whorled, and vine-like archetypes. Preliminary testing showed UniRig collapsing complex branching into near-linear chains, while SkinTokens preserved topology better but over-segmented branches and produced an unstable output space, so we focused on UniRig for its greater stability. Diagnosis traced the collapse to sampling-level suppression of branch tokens, and further analysis showed the frozen mesh encoder had limited sensitivity to structural variation, pointing to a geometric bottleneck in the tokenization pipeline rather than a purely learned bias. Building on these findings, we applied multi-round fine-tuning over multiple procedurally generated synthetic datasets. Across rounds, the model progressively recovered accurate branching topology and generalized beyond branch-only structures to plants with foliage, a harder case given the zero-thickness, mesh-normal-dependent geometry of leaves. The resulting model generalized well across diverse plant forms without leaf-specific architectural changes, indicating that targeted fine-tuning can substantially close the domain gap between character-rigging priors and plant skeletal structure. As such, our work points toward a viable path for automated plant rigging across both branch topology and foliage type, even those not considered in our findings.
Authors: Shengchao Chen, Ting Shu
Abstract: Federated learning enables medical-imaging models to be trained across hospitals, and privacy law, most explicitly the GDPR ``right to be forgotten'', turns removing a hospital's, a class's, or a patient's influence from such a model into a federated unlearning problem. This need is most acute in medicine, where patients withdraw consent and hospitals leave collaborations. Yet nearly all unlearning evidence comes from natural images, whose heterogeneity and task structure differ sharply from clinical data, so it is unclear whether existing methods transfer, and no shared protocol covers clinical data. We present Lethe, a benchmark for federated unlearning in medical imaging. It evaluates twelve methods across eight task families, from classification and segmentation to denoising, cross-modality synthesis, and vision-language question answering, at three forgetting granularities and against a retrained gold standard on utility, privacy, and cost. The central result is that what separates methods is the difficulty of the forgetting request, not the method itself. The easy removals that dominate the literature leave the methods that preserve utility indistinguishable, while only hard ones separate them. More striking, on the many medical tasks that generalize across sites, forgetting a client barely changes task performance, leaving residual membership as the signal that must be erased.
Authors: Xiaoqian Lu, Guangfu Guo
Abstract: Object-centric models often produce fragmented masks, boundary leakage, and incorrect region merging. We introduce Similarity-Shift Refinement (SSR), a training-free post-hoc method for improving object-centric masks with a frozen self-supervised Vision Transformer. SSR measures changes in pairwise patch similarity before and after self-attention value aggregation, retains positively strengthened relations, and constructs a sparse affinity graph. This graph propagates the initial soft slot assignments in a single refinement step, without retraining or modifying either model. Across natural-image, synthetic-video, and real-world-video benchmarks, SSR improves all-pixel Adjusted Rand Index in all 24 evaluated model-dataset combinations, with an average gain of 8.5 percentage points. Ablations show that value-space similarity shifts outperform query- and key-space variants as well as static Transformer affinities. However, texture-dense scenes may cause visually similar regions to be over-grouped. Overall, SSR provides a simple and transferable signal for training-free object-centric mask refinement.
Authors: Di Zhang, Li Zhang, Jiashuai Liu, Junbo Lu, Zhi Zeng, Jiusong Ge, Chunze Yang, Yi Niu, Jian Chen, Kai He, Zeyu Gao, Chen Li
Abstract: Whole slide image (WSI) classification is an evidence-driven task, where diagnostic cues are often sparse, spatially organized, and class-dependent. Existing MIL and vision-language methods aggregate a large pool of patch features into a single global slide representation. Under few-shot supervision, limited slide-level labels make it difficult to learn a reliable aggregation mechanism that organizes sparse local cues into compact and coherent diagnostic evidence. Moreover, a shared slide representation compresses evidence supporting a candidate class and its alternatives into the same feature, limiting class-specific reasoning and interpretability. To address these issues, we propose EviBall, a class-conditioned evidence retrieval framework for few-shot WSI classification. EviBall organizes local patches into Evidence Balls through semantic-spatial assignment and center refinement, yielding compact and spatially coherent evidence units under weak supervision. It then uses task-specific class queries, including language-guided queries for morphology-oriented tasks and molecular-guided queries for molecular endpoint prediction, to retrieve supporting evidence balls and produce class-conditioned evidence representations for direct class-wise prediction. By introducing structured evidence units and task-relevant semantic guidance, EviBall reduces the reliance on learning an unconstrained global aggregation mechanism from scarce slide-level labels. It therefore reformulates few-shot WSI classification as structured evidence retrieval and competition among candidate classes. Extensive experiments across four morphology-oriented and molecular endpoint WSI tasks demonstrate that EviBall consistently outperforms conventional and vision-language MIL baselines under diverse few-shot settings, while providing spatially localized and class-specific evidence for each prediction.
Authors: Junsheng Wang, Chao Chen, Mengying Xie, Mingyan Li, Fuqiang Gu
Abstract: Understanding and generating spatially coherent layouts from natural language remains a fundamental yet challenging task for large language models (LLMs). Existing LLMs often struggle to capture explicit geometric relationships and structural dependencies between objects. To address this issue, we propose SG-Layout, a graph-guided layout generation framework that explicitly incorporates structured spatial knowledge into LLMs. SG-Layout follows a two-stage training paradigm: (1) a graph-language feature alignment stage, where a relational graph encoder and a projector are trained to map scene-graph embeddings into the LLM's linguistic space; and (2) an instruction tuning stage, where LoRA-based adapters enable efficient fine-tuning for instruction-driven layout generation while keeping the backbone frozen. We evaluate SG-Layout on image layout generation, indoor scene synthesis and robotic object rearrangement tasks. Experimental results show that SG-Layout improves spatial reasoning accuracy and geometric consistency over the compact open-source backbone, with particularly clear advantages in relation-dense and compositionally complex scenes. These results highlight the effectiveness of graph-structured feature alignment for enhancing controllable layout generation.
Authors: Sen Liang, Fengbin Guan, Youliang Zhang, Xin Li, Zhibo Chen
Abstract: Text-driven instruction-based video editing in complex scenes remains challenging: purely textual prompts often fail to capture precise spatial relationships and physical constraints, resulting in target ambiguity and physically implausible outcomes. To address this, we propose a plan--guide--edit framework that explicitly bridges semantic intent and spatial execution. In our framework, a Chain-of-Thought (CoT)-enhanced multimodal large language model (MLLM) serves as a planner, performing structured reasoning over the video and instructions to derive a precise sequence of bounding boxes and attribute-enriched editing directives. These spatial priors then guide a box-conditioned mask generator, transforming ambiguous global retrieval into localized, context-aware refinement and producing masks that more accurately capture object scale, contact relationships, and placement. Building on these spatial and semantic signals, a diffusion-based editor integrates the masks, enriched instructions, and frame features to render high-fidelity edits that remain temporally coherent and spatially well aligned. Trained first in a modular manner and then jointly, our framework achieves superior performance with reduced data requirements, delivering precise localization in scenes with multiple similar objects and physically consistent object additions, and extensive experiments demonstrate state-of-the-art performance over multiple strong baseline methods. More details are available at: https://github.com/flying-sky999/CoT-Edit
Authors: Yian Zhao, Ruochong Zheng, Hongcan Guo, Yu Yan, Jian Zhang, Jie Chen
Abstract: Video world models predict future observations conditioned on historical observations and control signals, enabling long-horizon generation through autoregressive state transitions. Unlike conventional video generation models that primarily capture visual appearance and motion, video world models learn the underlying dynamics governing environment evolution under agent actions, providing a foundation for embodied AI and interactive simulation. Recent progress has largely relied on adapting pretrained video generation models through post-training or distillation. Although effective, these approaches often require complex training pipelines, substantial computational resources, and suffer from the mismatch between bidirectional pretraining and causal streaming inference. Recent studies have shown that training autoregressive video world models from scratch is feasible and scalable. However, the community still lacks a lightweight, transparent, and fully reproducible baseline trainable end-to-end with modest computational resources. We present MiniWorld, a reproducible framework for training streaming video world models from scratch. MiniWorld employs a block-causal Video Diffusion Transformer trained with Flow Matching in the latent space of a pretrained Video VAE. Building on Diffusion Forcing, it adopts a chunk-wise non-decreasing noise schedule and two-stage continued training to improve temporal modeling and stability. During inference, MiniWorld combines a rolling KV cache with pipelined asynchronous denoising for efficient streaming generation under bounded computation. The entire model can be trained within several days on a single 8-GPU server. By releasing the training and inference codebase and pretrained checkpoints, we hope MiniWorld will facilitate future research on video world modeling.
Authors: Chi Zhang, Haoyang Shi, Yueyi Liu, Zhaokun Yan, Yishu Yin, Yuhang Wu, Miao Liu
Abstract: Large language models have made text the default medium for human--AI interaction, buttext alone cannot express the full range of responses required by multimodal assistants,avatars, and embodied agents. While recent audio-video generative models can synthesizehigh-fidelity synchronized content, existing supervision is largely \emph{descriptive}:models are trained to render captions rather than to produce audio-visual responsescaused by external user interactions. We introduce \textbf{InteracVid}, \emph{the firstopen-source large-scale dataset that addresses this missing supervision}, so that everysample couples a preceding audio-visual context and an external stimulus with the realinteractive response that follows. We design a metadata-aware pipeline that extractsinteractive clips from long, noisy livestreams, yielding over \textbf{454K}context-query-response triplets from more than \textbf{59K} livestream videos andspanning conversation-centered, object-centric, procedural, embodied, and screen-basedscenarios. A ten-rater human study confirms that the extracted interactions are causal,natural, and temporally complete for both genuine and reconstructed queries. On aheld-out benchmark of \textbf{100} genuine live-chat queries, fine-tuning on InteracVidimproves both interaction planning and audio-video response generation, and anindependent human evaluation reproduces the system ranking and the conclusions obtainedwith our automatic judge. These results highlight interaction-structured data as acritical foundation for interactive multimodal generation.
Authors: Moxu Duan, Jingwen Fu, Yuwang Wang
Abstract: Streaming VideoLLMs process frames causally while visual tokens grow continuously, making compression essential for controlling prefilling latency and memory. Existing training-free methods independently rank tokens, ignoring marginal-gain interactions among retained tokens. We argue that streaming video token compression should instead be formulated as set selection, where each candidate is valued by what it adds beyond the tokens already retained. Unlike existing set-wise methods designed for offline tasks, streaming makes causal, frame-by-frame pruning decisions, so modeling cross-frame interactions requires an explicit historical reference. This creates a reference-set dilemma: the reference must adequately represent previously conveyed content while remaining bounded for real-time inference. We introduce NovaCov, to our knowledge the first training-free, plug-and-play set-wise token compressor designed for streaming video. NovaCov maintains a capacity-bounded, recency-weighted Historical Reference Bank and optimizes a dual-branch submodular coverage objective that preserves representative current-frame content while prioritizing information insufficiently covered by history. Both branches are facility-location functions, so greedy selection retains the classical (1-1/e) approximation guarantee. Across streaming and offline benchmarks, NovaCov outperforms existing training-free compression methods, retaining 99.6% of ReKV accuracy while reducing LLM prefilling latency by 46%.
Authors: Hongbo Duan, Pengting Luo, Chengzhi Zhao, Yuanhao Chiang, Fangming Liu, Xueqian Wang
Abstract: We present DynActiveGS, a dynamic-aware active reconstruction framework based on 3D Gaussian Splatting (3DGS) for autonomous exploration in dynamic environments. The framework incrementally reconstructs a 3D Gaussian scene representation while suppressing motion-corrupted observations through online uncertainty prediction and uncertainty-weighted Gaussian optimization. A key component of DynActiveGS is the explicit decomposition of uncertainty into structural uncertainty and motion-induced uncertainty, which enables the system to distinguish under-reconstructed static regions from dynamically unreliable areas. Based on these uncertainty fields, DynActiveGS performs dynamic-aware viewpoint selection and dynamic-constrained path planning to favor informative yet stable observations during exploration. The resulting system forms a unified closed-loop pipeline for robust active reconstruction in dynamic scenes. Extensive experiments on challenging dynamic benchmarks demonstrate consistent improvements over existing active reconstruction baselines in reconstruction accuracy, completeness, rendering quality, and exploration efficiency.
Authors: Changwoo Baek, Kyeongbo Kong
Abstract: Recent 3D vision-language models (3D VLMs) construct geometry aware tokens by projecting 2D visual features into world coordinates, enabling spatial reasoning for tasks such as 3D question answering. However, this design generates thousands of tokens per scene, resulting in substantial computational and memory overhead. While token compression has been extensively studied in 2D VLMs, existing approaches rely on semantic relevance or attention-based selection that overlook the structured spatial nature of 3D tokens. Moreover, redundancy in 3D representations cannot be resolved by spatial proximity alone, as object-level token imbalance persists even after spatial aggregation. To address this, we propose 3DZip, a three-stage token compression framework that first applies coarse voxelization to remove point-level redundancy, then selects anchor tokens based on feature-space diversity via a Determinantal Point Process, and finally merges remaining tokens under spatial constraints to preserve geometric coherence. Experiments on three 3D question answering benchmarks demonstrate that 3DZip consistently outperforms existing compression methods, retaining 94.7% of the original performance with only 128 tokens, achieving a $1.92\times$ faster inference speed.
Authors: Yinglong Li, Donghui Shen, Xiaoyu Zhang, Zhichao Ye, Hongyu Wu, Aimin Hao, Guofeng Zhang, Haomin Liu
Abstract: While feed-forward 3D Gaussian Splatting (3DGS) enables efficient 3D reconstruction, achieving high-fidelity rendering remains challenging. Existing pixel-aligned approaches suffer from spatial inflexibility and massive structural redundancy, whereas query-based methods lack 3D priors and entangle geometry with appearance, yielding blurry, pose-dependent results. To overcome these deficiencies, we propose \textbf{QuerySplat}, a feed-forward 3DGS framework driven by geometric priors and explicit appearance decoupling. Specifically, we design a dual-branch query-based decoder: the geometry branch leverages a pretrained Vision Geometric Model for spatial understanding, which intrinsically endows QuerySplat with pose-free modeling capabilities, while the appearance branch recovers high-frequency details through a dedicated pathway separated from geometric attribute regression. Extensive experiments demonstrate that QuerySplat mitigates the blurry rendering issues of early query-based models and consistently outperforms pixel-aligned approaches in rendering fidelity. On the challenging DL3DV benchmark, it achieves state-of-the-art novel view synthesis performance, with average PSNR gains of 2.30 dB and 1.04 dB over the best pose-free and pose-required baselines, respectively. Project Page: https://inspatio.github.io/querysplat.
Authors: Ahmed Baha Ben Jmaa, Faten Chaieb, Anna Fabija\'nska
Abstract: Fruit ripeness prediction (FRP) is a classification-based agricultural computer vision task that has attracted much attention, thanks to its wide-ranging advantages in agriculture field for both pre-harvest and post-harvest management. Accurate and timely FRP can be achieved using machine/deep learning-based hyperspectral image classification techniques. However, challenges including the limited availability of labeled data and the lack of robust methods generalizable to various hyperspectral cameras and fruit types can compromise the effectiveness of hyperspectral image-based FRP. Addressing these challenges, this paper introduces Fruit-HSNet, a machine learning architecture specifically designed for hyperspectral classification of fruit ripeness. Fruit-HSNet incorporates a spatio-spectral feature extraction module based on Fourier Transform and central pixel spectral signature followed by learnable feature fusion and a classifier optimized for ripeness classification. The proposed architecture was evaluated using the DeepHS Fruit dataset, the largest publicly available labeled real-world hyperspectral dataset for predicting fruit ripeness, which includes five different types of fruits-avocado, kiwi, mango, kaki, and papaya-captured with three distinct hyperspectral cameras at various stages of ripeness. Experimental results highlight that Fruit-HSNet substantially outperforms existing deep learning methods, from baseline to state-of-the-art models, with improvements of 12%, achieving a new state-of-the-art overall accuracy of 70.73%.
Authors: Puzhuo Zheng, Hasan Kurban
Abstract: Test-time scaling lifts large language model reasoning by sampling many candidate solutions and selecting among them, yet the same recipe transfers poorly to vision-language models (VLMs): recent work shows that simple majority voting beats selection methods built on the model's own self-verification, apparently because at the selection layer an image-grounded answer and a confident guess from the language prior look the same. A natural fix is to make the selection signal one that cannot be computed without the image. We study Perturbation Grounded Selection (Pgs), a label-free, training-free rule that scores each candidate by whether the model re-derives it under label-preserving perturbations of the input (cropping, background masking, mild photometric or geometric jitter); Pgs recovers majority voting when the perturbation set is empty. The decisive question is not whether Pgs beats chain-of-thought only majority voting, but whether the perturbation term adds anything once decoding format and budget are controlled. We therefore introduce a format-matched control (MatchedCtrl): the same short, no-CoT draws spent on the original image. Across TextVQA, MATH-Vision, MMMU, and ViLP, with a Qwen headline (three-seed means) and LLaVA-OneVision coverage in matched-budget selector tables, Pgs appears to beat plain majority voting by up to +31.8 points on TextVQA (Qwen), but MatchedCtrl tracks or exceeds Pgs within noise on every benchmark, including the vision-required ViLP; no Qwen category shows a significant gain over this control. The stability gap is real and image-dependent (up to +0.48), yet does not predict per-instance wins. The result is negative and diagnostic: perturbation consistency is at best a partial diagnostic of visual dependence and, on its own, not a usable selection signal once format is controlled; gains reported against CoT-only majority voting overstate such methods.
Authors: Haocheng Wang, Tongkun Guan, Wei Shen, Xiaokang Yang
Abstract: Visual document retrieval has recently become increasingly important in applications such as enterprise search, scientific literature discovery, and retrieval-augmented generation. These applications depend on efficiently identifying query-relevant pages across large collections of visually rich documents. Existing methods commonly adopt late-interaction architectures that encode and index documents offline to enable scalable and low-latency online retrieval. Despite its efficiency, this paradigm requires each document to be encoded into a fixed representation before the query is known. However, the same content in a visual document may induce different interpretations depending on the query intent, which a fixed representation struggles to capture. Yet postponing document encoding until the query arrives would incur prohibitive online retrieval latency. To address this gap, we propose VaRS-Doc, a visual document retrieval framework that diversifies document representations by enabling the model to actively explore variant latent interpretations during document encoding, while preserving efficient late-interaction retrieval in which each query adaptively selects the best-fit representation. We further introduce a two-stage training strategy that encourages the model to capture complementary semantic interpretations and prevents it from falling back to train a single dominant representation. Experiments on visual document retrieval benchmarks show that VaRS-Doc achieves state-of-the-art retrieval performance, offering a practical solution to the mismatch between query-agnostic document encoding and query-specific retrieval needs. Code is available at https://github.com/bokufa/VaRS-Doc.
Authors: Clarissa Rui Min Ong, Elizabeth Tee Inn Loo, Kenneth Woon Hao Soh, William Rachmadi, Qiming Zheng, Hao Li
Abstract: Indonesia's relocation of its political and administrative capital from Jakarta to Ibu Kota Nusantara (IKN) has been framed around a Forest City vision, yet rapid construction within the Core Government Area (KIPP) raises concerns over land conversion, vegetation loss, and carbon stock decline. This study applies remote sensing techniques to systematically assess land use and vegetation cover change in KIPP from 2021 to 2026 using PlanetScope SuperDove satellite imagery. Cloud-free mosaics were prepared and analysed through spectral indices, including the Normalised Difference Vegetation Index (NDVI), Normalised Difference Red Edge (NDRE), and Normalised Difference Water Index (NDWI), alongside supervised land use and land cover (LULC) classification using a Support Vector Machine algorithm. Results show substantial environmental transformation, with mean NDVI declined by 17.1%, total carbon stock decreased by 0.28%, developed land expanded by 672%, and total vegetation declined by 18.1%. Vegetation loss was most extensive between 2023 and 2024, although a temporary recovery in NDVI and carbon stock occurred from 2024 to 2025 as active clearing slowed and development shifted towards already-cleared land. Overall, the findings demonstrate that remote sensing provides an effective approach for monitoring the environmental impacts of large-scale urban development, while highlighting the need for higher-resolution, hyperspectral, and SAR-based methods to improve detection of construction stages and plantation-related land cover changes.
Authors: Zirui Zhang, Yinbo Yu, Donghai Guan, Chunwei Tian, Daoqiang Zhang, Qi Zhu
Abstract: The realism of images generated by multimodal large language models (MLLMs), such as GPT Image2 and Nano Banana2, has improved rapidly in recent years. Compared with early generative models, current models have made clear progress in text rendering. They can produce high-quality images that closely resemble real-world application scenarios. The enhanced generation capabilities of current MLLMs pose increasingly severe challenges to AI-generated image detection. Detection is no longer limited to identifying obvious artifacts left by early generators. Instead, it requires systematic and realistic benchmarks for the new generation of generated content. However, most existing benchmarks are still built around early generative models and cannot fully evaluate the forensic challenges introduced by high-quality and multi-form generated images. To address this gap, this paper constructs a benchmark dataset for detecting images generated by MLLMs. The benchmark covers several realistic application scenarios and adopts three generation protocols to simulate direct generation, reference-based reconstruction, and local editing. Based on this benchmark, we evaluate detector degradation from traditional scenarios to MLLM-generated images and analyze false positive rates and false negative rates across three sample types, revealing the failure modes of existing methods. We further propose a structural-artifact-prior-guided dual-stream prompt framework (SAP-DSP) as a strong baseline. SAP-DSP uses dual-stream prompt learning and structure-aware routing fusion to improve representation learning. Extensive experiments show that the proposed benchmark exposes the performance degradation of existing detectors on high-quality generated images, while SAP-DSP achieves more stable detection results on this benchmark. Our code and dataset are publicly available at https://github.com/xbrainnet/SAP-DSP.
Authors: Jiayang He, Tianling Xu, Diancheng Kang, Huaide Jiang, Junyan Bai, Shaoming Zheng, Xuan Song
Abstract: Video large language models (Video-LLMs) represent videos as dense sequences of visual tokens, whose length grows with the temporal and spatial extent of the input. These tokens often contain substantial redundancy arising from repeated visual patterns, leading to unnecessary computation in the subsequent language-model processing. Existing token compression methods, including pruning and merging, perform compression online during inference, repeatedly incurring additional computation for each input video and often relying on model-specific designs that limit their generality, we instead rethink this paradigm by shifting the costly compression process offline. We propose \textbf{ONCE}, a plug-in video token compression framework that introduces an offline-to-online paradigm: a frequency-aware global codebook is learned once in the visual feature space and reused for lightweight online compression through codebook lookup and aggregation, reducing repeated per-video computation and the need for model-specific compression designs. Extensive experiments across multiple video understanding benchmarks and against diverse compression baselines demonstrate that our approach achieves a strong accuracy-efficiency trade-off, maintaining competitive performance while achieving the lowest inference latency among compared methods.
Authors: Shuliang Zhu, Qi Wang, Ryugo Morita, Jinjia Zhou
Abstract: Full-body capture from unconstrained photographs requires global correspondence across arbitrary views, poses, crops, and occlusions. Yet pose, geometry, and foundation features estimated in this setting are too unreliable for dense matching or appearance transfer, while diffusion rectifiers and optimization pipelines expose no common interface for consuming such uncertain correspondence. Our insight is that correspondence need not be locally accurate: its coarse viewpoint and body layout can still organize how a diffusion prior adapts and guides reconstruction. We introduce \emph{Astrolabe}, a host-portable adapter built on frozen viewpoint-guided spherical maps (SPH). A fixed bounded transform converts SPH into a spatial noise shift, which is matched during prior adaptation and reused during downstream denoising or score-distillation guidance in both pipeline categories. When a rectifier exposes a reference router, the same target/reference SPH additionally supplies coarse compatibility scores to select native appearance features; router-free optimization uses only the shared shift path. Astrolabe therefore follows one SPH--shift--adapt--guide process without dense warping or a learned control branch. Across Puzzle-IOI and 4D-Dress, it improves all reported image metrics in both hosts and all paired Puzzle-IOI geometry metrics; image gains extend to rear views, while 4D-Dress geometry remains stable overall.
Authors: Jiawei Guo, Junxian Li, Yixin Tang, Bingya Zhang, Jiaxin Lu, Yulun Zhang, Shangchen Zhou
Abstract: Recently, diffusion-based removal methods have achieved promising visual quality in removing both target objects and their associated effects. However, they typically rely on multi-step denoising, leading to high inference cost. Directly applying existing one-step distillation methods is also suboptimal, since their global objectives lack explicit region-wise calibration and may weaken the asymmetric edit-and-preserve behavior required by object-effect removal. To address these challenges, we propose TurboClear, a one-step SDXL-based object-effect removal model. During training, we design Region-Calibrated Distribution Matching (RDM) for region-aware distillation to preserve the teacher model's asymmetric edit-and-preserve behavior. Furthermore, we propose Learnable Spatial Fusion (LSF) for lightweight inference-time fusion. Extensive experiments show that TurboClear significantly improves inference efficiency while maintaining competitive visual quality. TurboClear reduces the computational overhead by up to $40.04\times$ compared to ObjectClear, and by up to $665\times$ against the Flux-based method OmniPaint, all while maintaining comparable or better visual removal quality. Code is available at https://github.com/GuoCalix/TurboClear.
Authors: Junno Yun, Ya\c{s}ar Utku Al\c{c}alar, Mehmet Ak\c{c}akaya
Abstract: Diffusion Transformers (DiTs) have emerged as a core architecture in generative modeling due to their scalability and adaptability to multimodal tasks. DiTs comprise isotropic transformer blocks, and learn representations progressively across depth, where the denoising objective drives later layers to focus on fine-detail reconstruction. This results in degraded representation quality and an imbalanced encoder-decoder behavior. Prior approaches such as representation alignment (REPA) mitigate this by encouraging stronger early representations via training regularization. Alternatively, U-Net-style DiT architectures introduce explicit multi-scale encoder-decoder structures for improved convergence. But they build on standard U-Net wisdom via learnable operators for spatial downsampling, which are not well-suited to transformer architectures, introducing inefficiencies and compatibility issues with components such as cross-attention and representation regularization. In this work, we propose UDT, a U-Net diffusion transformer that combines the representation power of DiTs with the encoding-decoding benefits of U-Nets, through data-adaptive token merging for downsampling and upsampling, while preserving the DiT token dimension. Our baseline UDT architecture outperforms existing U-Net DiTs and achieves performance comparable to REPA across all model sizes. Furthermore, using architectural optimization and REPA, UDT outperforms SiT's 7.9 FID at 1400 epochs (w/o CFG) within 40 epochs (~ 40x faster convergence) for XL model size on 256x256 ImageNet. Finally, it achieves strong image generation performance with CFG, reaching FID of 1.38 (320 epochs) with SD-VAE and 1.35 (500 epochs) with VA-VAE, providing a new backbone for DiTs with strong empirical benefits.
Authors: Haoran Liu, Mingzhe Liu, Peng Li, Guibin Zan
Abstract: Infrared-visible image fusion (IVIF) has no ideal fused reference, so fusion algorithms are routinely ranked by scalar objective metrics that formalize different proxies for information transfer, structure, or source similarity. These proxies often disagree with the judgment that ultimately matters: given the same sources, which of two fused results does a human prefer? Direct pairwise comparison is an established reference protocol for relative subjective assessment, but its cost grows quadratically with the number of algorithms, which prevents routine use. We present the Learned Perceptual Image Fusion Measure (LPIFM), a source-conditioned model that operationalizes the human A/B/Tie comparison protocol as a repeatable, scalable surrogate. LPIFM jointly observes the infrared source, the visible source, and two fused candidates, and predicts whether candidate A is better, candidate B is better, or the two are perceptually equivalent. Supervision comes from a new dense preference corpus that covers every unordered comparison among a broad pool of fusion methods on the scenes of a public benchmark, labeled under a blinded, randomized, two-stage protocol with expert adjudication. Across scene- and method-generalization settings, LPIFM tracks human pairwise decisions closely and reproduces the tie-aware Bradley-Terry rankings derived from human labels; on full method pools it surpasses the strongest conventional metric by a wide margin in both pairwise accuracy and ranking correlation. We release the annotated preference dataset, together with the LPIFM model weights, source code, and evaluation code, to support preference-aligned IVIF assessment. LPIFM offers a practical instrument for human-aligned method comparison and ranking at scale.
Authors: Ziheng Wang, Chaolang Li, Yutong Yang, Xiaohan Xu, Chongxiang Yang, Hengxuan Zhong, Zhen Liang, Pengwen Dai
Abstract: State-of-the-art RGB-Event detectors improve the detection of small, fast-moving objects by combining complementary features from RGB and Event data, yet they typically fuse the two modalities into a unified representation for both localization and classification. Such a task-symmetric design is inconsistent with the intuition that the two modalities should play different roles according to their task-specific strengths. To examine this issue, we conduct a modality-specific evaluation and find that the relative advantage of the two modalities reverses across tasks: Event data are substantially more effective for class-agnostic localization, whereas RGB data provide stronger category evidence within localized target regions. Motivated by this task-dependent asymmetry, we propose an Asymmetric Event-RGB Object Detection Transformer (AERODet). During class-agnostic localization, Scale-wise Uncertainty-aware Reliability Estimation (SURE) calculates the relative reliability of the two modalities from their objectness response heatmaps and accordingly calibrates their contributions when the decoder aggregates multimodal features. Once the candidate boxes are obtained, Task-Decoupled Semantic Refinement (TDSR) decouples classification from localization and uses RGB RoI features for fine-grained classification. Extensive experiments on FRED and NeRDD demonstrate that AERODet achieves state-of-the-art performance. In particular, it surpasses the strongest RGB-Event baseline by 10.7 mAP points on the FRED challenging split.
Authors: Yibin Huang, Jixiang Hong, Zongzhao Li, Yuhan Dai, Zhibin Wang, Chunwei Wang, Jun Song, Chen Wang, Xiaofei Sun, Xiaoxiao Xu, Conghui Zhu
Abstract: Latents from vision foundation models (VFMs) are semantically rich and well suited for visual understanding. Recent representation autoencoder methods such as RAE have shown that they can provide promising latent spaces for image generation. However, VFM latents remain difficult to model directly: DiT-generated latents exhibit spectral mismatch with encoder latents, especially in high-frequency components. Our channel-wise spectral analysis further reveals that these high-frequency components are diffusely distributed across latent channels and entangled with semantic information, making the latent space difficult for DiT to model. To address these challenges, we propose SPAE, latent adaptation framework for generation. Specifically, SPAE employs a compact bottleneck to distill stable semantic information while suppressing high-frequency components, thereby improving the alignment between DiT-generated latents and encoder latents. In addition, we apply a channel-wise masking strategy to promote the decoupling of semantic information and high-frequency details across bottleneck channels. Experiments show that SPAE achieves a favorable balance among visual understanding, generation quality, and reconstruction fidelity.
Authors: Jianmin Chen, Jiaqi Tang, Wei Wei, Xiaogang Xu, Jiafei Wu, Zhe Liu, Qianzhou Wang, Yingying Yan, Botong Geng, Yuyang Xia, Lei Zhang, Qifeng Chen
Abstract: Multimodal large language models (MLLMs) increasingly rely on long chain-of-thought reasoning for complex tasks. However, as reasoning sequences lengthen, models may gradually rely less on visual evidence and more on accumulated textual context, leading to visual forgetting. Existing approaches do not directly constrain how visual evidence is used and maintained along the original reasoning trajectory, leaving long-context visual forgetting insufficiently addressed. To address this issue, we propose Remember-R1, a reinforcement learning framework that mitigates long-context visual forgetting by applying process-level supervision directly on the original reasoning trajectory. Specifically, Remember-R1 introduces rewards that encourage broader coverage of matched visual keywords, stronger persistence of visual dependence in later reasoning steps, and greater focus on question-relevant image regions. Experiments across multiple model scales and diverse multimodal benchmarks demonstrate that Remember-R1 consistently improves reasoning performance. Additional analyses further show that it slows the decline of visual attention during generation, supporting its effectiveness in mitigating long-context visual forgetting.
Authors: Fei Li, Yue Yu, Yuran Wang, Xinghan Li, Jingjing Chen, Yu-Gang Jiang
Abstract: AI-generated video (AIGV) detection aims to distinguish real videos from AI-generated ones. In practice, detectors trained on existing data often fail to generalize to newly emerging generative models, making this task challenging. Therefore, continual learning (CL) is essential for improving the adaptability. However, CL frameworks for this task remain underexplored. To this end, we propose SphereVideo, a novel CL framework for AIGV detection built on two key observations. First, real videos exhibit a compact feature distribution. Based on this, we encourage real video features to cluster around a real prototype on a hypersphere while repelling AI-generated samples, thereby establishing a decision boundary. This prototype serves as a stable anchor for CL, regulating boundary evolution and mitigating catastrophic forgetting. Second, existing methods tend to rely solely on spatial artifacts as shortcuts. To enhance temporal modeling, we introduce a strategy that models the temporal dynamics of real data at both frame and clip levels. By strengthening real data modeling, this strategy further facilitates learning a real prototype and forming a stable decision boundary. Moreover, we construct a comprehensive and challenging benchmark. Extensive experiments demonstrate that SphereVideo achieves an improved plasticity-stability trade-off, outperforming prior methods by 3.08% on seen data and 4.00% on unseen AI-generated data.
Authors: Advait Pavuluri, Shamik Karkhanis, Uzma Mushtaque
Abstract: Modern autonomous-driving fleets record far more video than human reviewers can inspect. This motivates the need for an automatic clip triage mechanism, to surface rare and review-worthy clips, so that driving models can be fine-tuned to better handle unideal circumstances. We test a label-free approach that scores clips by the prediction-error "novelty" of a self-supervised joint-embedding predictive architecture (JEPA); a frozen V-JEPA video encoder is paired with a lightweight predictor head to reconstruct masked clip embeddings, and clips whose embeddings are hard to predict are flagged as interesting. Evaluated under a realistic protocol that trains on one dataset and tests against footage from others, this approach appears highly effective. We show that this apparent success is actually a domain-shift consequence: on a fair benchmark drawn from a single dataset, this mechanism collapses to chance and is on par with simple no-training baselines. A lightly supervised probe on the same frozen embeddings results in almost double the average precision, indicating that the bottleneck is indeed the self-supervised objective, rather than the representation. We present this as a study for evaluating the effectiveness of self-supervised learning, where cross-dataset protocols can silently reward domain separation over novelty.
Authors: Pan Yin, Runtian Xia, Weisong Kuang, Kaiyu Li, Cong Zhao, Xiangyong Cao
Abstract: High-definition (HD) maps are essential for autonomous driving systems. In constructing such maps, onboard multi-view camera images, standard-definition maps and satellite images provide crucial information. However, due to the modality and perspective differences among these data sources, existing methods often struggle to effectively align and fuse them, making online HD map construction still challenging. To address these issues, we propose Driver2Map, an online HD map construction model inspired by human drivers. Unlike existing HD map construction models that utilize only two modalities, our Driver2Map can simultaneously exploit three modalities. Specifically, we propose a "two-stage alignment" strategy to reduce spatial misalignment across different modalities. Additionally, we introduce "Pose-Guided BEV Fusion", a BEV (bird's-eye-view) generation module that leverages camera pose information to adaptively weight multi-view features, thereby effectively suppressing cross-view feature overlap during BEV generation. Also, we design a "Pretrained Prior for Map Refinement" module to refine the initial prediction by learning map structure priors, thus improving the HD map prediction under dynamic occlusions. Extensive experiments demonstrate that Driver2Map outperforms existing methods on both IoU and AP metrics.
Authors: Reyhaneh Hosseinzadeh, Parham Zilouchian Moghaddam, Mehdi Modarressi
Abstract: The emergence of transformer-based deep learning models has brought unprecedented performance across various domains, particularly in natural language processing and computer vision. However, deploying these models, especially on resource-constrained devices, poses significant challenges due to their high computational complexity and large memory size and bandwidth requirements. This complexity has led researchers to use low-bit model weights to reduce memory usage and improve efficiency. In addition to reducing processing and memory demands, quantization introduces another useful property: value locality, where the extremely large number of parameters are restricted to a limited range of values. To fully take advantage of this locality, this paper presents DeVIT, an acceleration method for vision transformers that leverages differential computation to enable multiplier-less matrix multiplication.
Authors: Linhai Zhuo, Junxi Cai, Tianwen Qian, Qingping Zheng, Yang Liu
Abstract: Data augmentation, which simulates diverse visual variations to expand the source distribution and induce synthetic domain shifts, is a simple yet effective strategy for mitigating severe domain shifts and limited labeled target data in cross-domain few-shot object detection (CD-FSOD). Existing approaches rely on conventional data augmentation, such as Color-Jitter, Mosaic, and background-centric adaptation (e.g., Domain-RAG), which are limited in modeling complex domain shifts and often lead to suboptimal performance. In this paper, we propose PSP-FSOD, a principled framework that integrates prompt-driven domain simulation with feature perturbation regularization to improve generalization in CD-FSOD. To enable controllable domain synthesis, we design a prompt-driven strategy that leverages the visual grounding capability of large VLMs to jointly model foreground and background variations, generating semantically consistent yet domain-diverse training samples. Moreover, we adopt a grounding-aware generation scheme that guides object placement and alleviates semantic-spatial misalignment, thereby improving foreground adaptation. To ensure training stability and robustness, we further introduce a noise-induced feature perturbation mechanism that injects Gaussian noise into multi-scale intermediate features with distribution correction, encouraging consistent predictions under perturbations and reducing reliance on domain-specific cues. Extensive experiments demonstrate that PSP-FSOD produces high-quality domain-diverse supervision and learns domain-invariant representations, consistently improving performance across CD-FSOD benchmarks.
Authors: Yicheng Xiao, Haoxuan Ma, Caorui Li, Yucheng Wu, Weijie Wang, Haoxiao Wang, Shuang Chen, Fan Yang, Haiyun Guo, Jinqiao Wang
Abstract: Recent studies develop pixel-level multimodal large language models (MLLMs) that support both Region Segmentation and Region Understanding, extending multimodal interaction from whole images to specific objects and regions. However, these methods face two fundamental challenges. First, the scarcity of high-quality mask--text pairs leaves abundant mask annotations without corresponding language supervision. Second, discrepancies in supervision formats and learning-signal densities induce optimization interference between Region Segmentation and Region Understanding. To address these challenges, we propose PixVL, a self-supervised post-training framework that introduces a unified Mask--Text Consistency Cycle, enabling pixel-level MLLMs to generate and self-verify regional descriptions and learn from unlabeled data. We found that direct cycle based solely on geometric reconstruction is unreliable because re-segmentation IoU does not faithfully reflect the semantic quality and referring sufficiency. PixVL therefore introduces confuser-aware semantic verification, which uses the model's confidence when it correctly chooses the target among highly similar candidate regions, and assigns zero reward to an incorrect choice. Meanwhile, PixVL performs cross-view verification using temporally separated video frames or geometrically transformed image views, preventing cyclic learning from collapsing to positional and shape shortcuts. Finally, a quality-coupled bidirectional learning strategy uses the highest-reward description to guide Text-to-Mask learning. This strategy transforms Region Understanding and Region Segmentation from competing tasks into mutual generators and verifiers. Experiments demonstrate that PixVL improves both region understanding task and segmentation task.
Authors: Junhan Wang, Kani Chen
Abstract: Decoding visual experience from non-invasive brain activity is central to neuroscience and brain-computer interfaces. Functional magnetic resonance imaging (fMRI) offers fine spatial detail, but its slow hemodynamics and burdensome acquisition limit temporally resolved decoding. Electroencephalography (EEG) and magnetoencephalography (MEG) provide millisecond resolution, making image retrieval compelling: identify the viewed image from one neural response and a fixed candidate bank. Contrastive alignment to pretrained visual representations enables zero-shot retrieval from EEG and MEG, but most systems collapse heterogeneous visual supervision into a single embedding before ranking. This early consolidation imposes one similarity geometry on every candidate order and removes encoder-specific disagreements from the final ranking. We propose CORTIVA, a candidate-score fusion framework that preserves this complementary evidence. Three decoding routes are aligned to heterogeneous visual targets, score the same indexed candidates independently, and combine only their temperature-scaled score vectors before ranking. On the 200-way THINGS-EEG2 benchmark, CORTIVA reaches 73.5% Top-1 and 95.3% Top-5 across ten participants, exceeding the strongest reported baseline by 10.3 and 5.4 percentage points. With a modality-specific neural encoder, the same fusion principle reaches 42.4% Top-1 on THINGS-MEG. Matched route-removal retraining and four weight controls demonstrate that CORTIVA's gain arises from integrating complementary route scores and persists with uniform weighting, without requiring a specialized weighting rule. Independent DINOv2 analyses further reproduce the local error neighborhoods and posterior neural-visual correspondence. These results establish candidate-score fusion as a simple and testable alternative to embedding-level consolidation for neural image retrieval.
Authors: Zhiwei Chen, Yang Hu, Yuxiang Xiao, Yakun Ju, Tianyang Zhang, Yingxue Xu, Wei Li, Hao Chen, Jens Rittscher, Kaixiang Yang
Abstract: Pathology foundation models (PFMs) provide strong tissue representations and have become central to digital pathology. However, deployment in disease-specific settings is limited by 1) the high computational cost of billion-parameter PFMs and 2) distribution mismatch and non-biological bias inherited from pan-cancer, multi-centre pre-training, including site-specific signatures and imbalanced disease prevalence. These factors can encourage shortcut learning and under-emphasise subtle morphology required for reliable modelling of a specific cancer type. We present SmartStu (a Smart Student), a framework to customise compact, breast-cancer-specific PFMs via distillation whilst mitigating confounding. SmartStu distils representations from multiple teacher PFMs into a lightweight student backbone. Crucially, we introduce adversarial distillation that leverages a dedicated noise model trained to predict nuisance, edge-dominated cues on the distillation set. Using this noise model as a counterexample, the adversarial objective encourages the student to recognise, yet suppress, features predictive of nuisance targets. We further incorporate multi-teacher ensemble distillation and an auxiliary self-supervised objective with artefact injection. We validate SmartStu on three external cohorts (Yale HER2, SLN-Breast, and BRACS) with multiple tiny backbones. SmartStu yields breast-cancer-specific PFMs that are over $30\times$ smaller than general PFMs whilst largely preserving, and sometimes improving, downstream performance measured by balanced accuracy (bAcc) and AUC. Code is available at https://github.com/zwchen03/advDistall.
Authors: Yuxiang Xiao, Yang Hu, Bin Li, Tianyang Zhang, Zexi Li, Huazhu Fu, Jens Rittscher, Kaixiang Yang
Abstract: Pathology foundation models (PFMs) provide strong tile-level representations via self-supervised pre-training on large-scale pathology images. Yet, PFMs are developed under diverse and often opaque data, architecture, and objective choices, inducing latent representational biases that limit robustness and obscure what each model specialises in. We present AdaFusion, a lightweight adaptive fusion framework that integrates complementary signals from multiple frozen PFMs through (1) low-dimensional feature compression and (2) a sample-conditioned gating module that reweights model-wise (and optionally channel-wise) contributions. Beyond improving predictive accuracy, AdaFusion provides contribution-driven interpretation that offers evidence consistent with model-specific preferences and synergistic interactions across tissue phenotypes. We evaluate AdaFusion on three public benchmarks spanning treatment response prediction, prostate cancer grading, and spatial gene expression inference. AdaFusion consistently outperforms individual PFMs and other fusion baselines, while providing interpretable tissue visualisation which aligns model preferences with morphological patterns. Code is available at: https://github.com/xyx-98/PathoOracle.
Authors: Tongyan Wang, Zhengyuan Li, Muhan Lin, Shengyang Luo, Yifan Shen, Aniket Bera, Baijian Yang, Yingjie Victor Chen
Abstract: Text-conditioned human motion generation has made rapid progress with the emergence of large-scale motion--language datasets. However, even datasets with rich long-form descriptions typically provide supervision only at the clip level, without explicit temporal correspondence between motion frames and language. This limits fine-grained motion--text grounding and temporally precise generation. We propose FineMoLA, a weakly supervised framework that learns fine-grained frame--phrase correspondence directly from clip-level annotations. Our method first segments long-form descriptions into action-bearing phrases, and then formulates motion--language alignment as an optimal transport problem, which naturally models many-to-many relations between motion frames and text under global constraints. With entropic regularization and Sinkhorn iterations, FineMoLA efficiently infers pseudo frame-level alignments without human labeling. Experiments on SnapMoGen demonstrate that the learned alignments outperform baselines in motion--text grounding.
Authors: Sabri Mustafa Kahya, Richard R. Chen, Muhammet Sami Yavuz, Jerry Jierui Lou, Akanimoh Adeleye, Haci Ali Kahya, Jana Lipkova
Abstract: Safe deployment of AI methods in medicine requires robust guardrails that detect when input data deviate from the training distribution to ensure that models provide predictions only within their scope of expertise and abstain otherwise. Out-of-distribution (OOD) detection can provide such safeguards and is extensively studied in general computer vision. Yet, it remains underdeveloped in computational pathology, where gigapixel whole-slide images (WSIs), subtle differences between disease subtypes, and variability in tissue preparation pose unique challenges for conventional OOD methods. We propose ZIO, a training-free, multimodal OOD detector for pathology WSIs that leverages vision--language pathology foundation models (FMs). ZIO constructs text and visual prototypes of in-distribution classes and integrates their complementary information through a prototype shrinkage mechanism to derive OOD scores. We provide the ZIO formulation for both slide- and patch-level FMs. We evaluate ZIO across diverse clinically relevant domain shifts, including rare diseases and near-OOD settings. Extensive evaluation of over 14,700 WSIs from five independent consortia shows that ZIO consistently outperforms both unimodal prototypes and 40 state-of-the-art OOD methods. These results demonstrate the benefits of multimodal representation for OOD detection and pave the way towards safer AI deployment in clinical practice.
Authors: Yi You, Hui Li
Abstract: 3D bin packing rectangular items into standardised containers to maximise space utilisation under geometric shipping automation. Loading a furniture purchase into a personal vehicle is the same task, but under more complex conditions that standard container loading algorithms ignore. This paper addresses the physically stable placement under these realistic conditions with heterogeneous boxes (e.g. varying dimensions and weights) and occupied containers (e.g. groceries). This paper provides a real-world benchmark dataset and baseline model for the Heterogeneous furniture-in-vehicle packing task. The dataset uses real furniture company flat-pack packaging data covering a large number of catalogue products via family-level extrapolation with diversity length, widths, heights, and weights. We also propose a PackingGPT framework for packing as a sequential placement inspired by the Lego assembly process, where heterogeneous boxes of varying dimensions (bricks) are placed step-by-step into the irregular remaining cargo space (creations). Five baseline packing methods were tested on our dataset without considering the Centre-of- Mass (CoM) constraints. In sedan car simulations, 10-40% of placed boxes failed the stability check on average. When the LLP model was trained on packing sequences with CoM constraints enforced during placement, the failure rate dropped to 0.67% (SUV-500).
Authors: Bingxuan Li, Rui Yang, Cheng Qian, Jiateng Liu, Jeonghwan Kim, Zhenhailong Wang, Manling Li, Tong Zhang, Heng Ji
Abstract: Agents are increasingly expected to act not only as task executors, but also as decision-makers on behalf of human users. This shift requires agents to accumulate evidence over long horizons, interpret implicit user preferences, and compare multiple candidates under partial observations. In this work, we propose DunphyBench, a new benchmark for evaluating agents on long-horizon human-centered embodied decision-making, where the agent must navigate through multiple embodied housing environments and make decisions that align with multi-dimensional human preferences. Unlike standard embodied reasoning tasks that often focus on procedural planning or immediate goal completion, our setting requires agents to integrate multimodal, multi-source input into coherent knowledge that supports complex reasoning across long horizon. The evaluation results reveal that there is a substantial gap between current agents and human performance. Furthermore, our diagnosis of state-of-the-art VLM-driven agents reveals that memory management is one of the bottlenecks, where raw multimodal history introduces noise that hinders decision quality. Motivated by this finding, we design MeMento, a preference-conditioned multimodal memory compressor that selectively compresses decision-relevant information from long-horizon history based on user preferences with a fixed set of memory tokens. Experiments show that MeMento helps VLM-driven agents improve accuracy by 7.18%, while reducing memory usage by 85.38% compared to the strongest baseline.
Authors: Youxin Jiang, Baoheng Fu, Hongwei Ren, Xiangqian Wu
Abstract: Event cameras produce sparse and asynchronous event streams that provide rich spatio-temporal information for efficient perception. Recent advances in event-based models have demonstrated strong performance by directly modeling asynchronous events without dense frame reconstruction. However, identifying the event-level evidence behind their predictions is crucial for improving model transparency and reliability. Directly adapting point-level saliency methods from point clouds provides fine-grained attribution but overlooks event-specific spatio-temporal structures. To address this limitation, we propose Voxel-Guided Global Event Ranking (VGER), a training-free attribution framework for point-based event cloud networks. VGER combines event-level gradient evidence with task-aware voxel perturbation evidence, transferring regional contribution into event-level attribution scores while preserving fine-grained resolution. Furthermore, VGER introduces a unified event ranking strategy, where high-ranked events are expected to be prediction-critical and low-ranked events are expected to have limited influence on predictions. We evaluate VGER on three event-based benchmarks with PointNet, PointNet++, and EventMamba. Across nine dataset-backbone settings, VGER consistently improves both high-tail and low-tail deletion performance over point-level saliency baselines.
Authors: Guiqiu Liao, Matjaz Jogan, Daniel A. Hashimoto
Abstract: Multimodal large language models (MLLM) for surgical scene understanding typically inject hundreds of dense visual tokens into a language model, leading to costly inference and limited spatial traceability for generated answers. We present Slot2Text, a dual-mode surgical MLLM that replaces dense representations of visual input with a compact set of regions encoded as slot latents. Instead of relying on contrastive alignment of the visual encoder with language, Slot2Text groups self-supervised vision features into a few regions--slots that are consumed by the language model as area-labeled visual tokens. Slot2Text-Fast uses the slot prefix to answer surgical questions. Slot2Text-Reason also identifies and locates areas relevant for reasoning, linking language outputs to corresponding slot tokens, masks or regions. Experiments on multiple visual question answering and visual grounding benchmarks show that Slot2Text-Fast is competitive with state-of-the-art baseline at a much lower cost, reducing the average total token consumption by a 91.8\% and the visual prefix from 1,295 to 47 tokens (a 96.4\% reduction). Slot2Text-Reason trades additional tokens and latency for explicit area identities, locations, and traceable spatial evidence. These results establish compact slot latents as an efficient default visual interface for surgical MLLMs, with grounded reasoning invoked when greater spatial traceability is required.
Authors: Yuqi Li, Yuedong Tan, Huiran Duan, Weilun Feng, Chuanguang Yang, Zhulin An, Zongwei Wu, Shiping Wen, Tingwen Huang, Yingli Tian
Abstract: Unified multimodal object tracking has achieved remarkable robustness by leveraging complementary sensor data (e.g., RGB, Thermal, Depth), yet the heavy computational burden of state-of-the-art models hinders their deployment on resource-constrained edge devices. In this work, we identify the prediction head as a critical but often overlooked efficiency bottleneck. By strategically streamlining the decoder architecture, we unlock the potential for real-time inference but simultaneously introduce a capacity gap between the lightweight student and the heavy teacher. To resolve this, we conduct a systematic analysis of 17 distillation strategies and introduce a Dual-Alignment Distillation framework. Our key insight is that effective compression requires decoupling knowledge transfer into two complementary streams: (1) Spatial Representation Alignment, which employs feature distillation to sharpen the student's spatial focus on foreground targets ("Where to track"); and (2) Semantic Distribution Alignment, which utilizes logit-based distillation to align decision boundaries and transfer discriminative dark knowledge ("What to track"). Extensive experiments across five benchmarks demonstrate that our approach significantly outperforms complex state-of-the-art methods. Notably, our distilled model achieves 91.5% MPR on RGBT234 and operates at 54 FPS on a single RTX 4090, representing a 5x speedup over the teacher model while maintaining superior accuracy.
Authors: Baihan Yang, Tiexin Li, Yuheng Liu, Xin Lin, Xinke Li, Xiaohui Xie, Truong Nguyen
Abstract: Selecting a complete 3D object from a reconstructed scene with minimal user effort is essential for practical scene editing and embodied interaction. Existing 3DGS-based methods either retrain the Gaussian representation to embed per-object labels, or build dense multi-view SAM observations, both requiring heavy computation and dense viewpoint coverage that is rarely available in practice. We present GaussianSelector, a training-free framework for interactive 3D object selection from sparse views and sparse scribble guidance. Operating directly on native Gaussian primitives, we coarsen dense Gaussians into geometrically coherent superpoints and construct a continuity-weighted graph using appearance and spatial cues. Sparse user scribbles are lifted into 3D via visibility-aware transmittance coverage, and selection is solved as a global graph-cut energy minimization that propagates sparse evidence to a complete 3D object. This design naturally supports multi-round refinement, where users iteratively correct the selection from additional viewpoints to progressively improve the result. Experiments demonstrate that GaussianSelector achieves competitive selection quality against state-of-the-art multi-view SAM-based methods, while requiring significantly fewer interaction views and substantially lower computational overhead. These properties make it well suited for human-in-the-loop 3D scene editing and 3D asset extraction in real-world deployment scenarios.
Authors: Robin Kim, Colin Samplawski, Benjamin M. Marlin
Abstract: Detection transformer models, including DETR and its extensions, learn to output a set of object-level embeddings that can be simultaneously decoded into 2D bounding boxes and class distributions. In this paper, we investigate what pre-trained 2D detection transformers understand about the 3D properties of objects. Specifically, we investigate the extent to which properties including the depth of objects from the camera and the 3D location of objects relative to the camera can be recovered from object-level embeddings using linear and non-linear probes. Across a range of detection transformer models, our results show a surprisingly strong and previously unknown ability of 2D DETR models to represent useful information about the 3D properties of objects, despite the complete lack of 3D supervision during model pre-training.
Authors: Yongcong Zhang, Navid Rabbani, Bangyan Liao, Chengbo Wang, Yizhen Lao, Adrien Bartoli
Abstract: Rolling shutter (RS) cameras are widely used in consumer devices, but their row-wise exposure causes distortions under motion, making geometric 3D vision problems dependent on both camera intrinsics and readout time ratio. Existing RS calibration methods rely on calibration targets or specialised hardware, limiting their use in unconstrained settings. We present the first self-calibration method for RS cameras that directly estimates camera intrinsics and the readout time ratio from image sequences, without requiring calibration targets. The method is implemented as a self-calibrating bundle adjustment (BA), which critically depends on the RS imaging model. We combine two known complementary models. The first formulates RS imaging as continuous-time trajectory estimation under a row-wise pose representation. The second interprets RS images as temporally distorted global shutter (GS) images and requires to estimate correction fields. The combination is non-trivial and results in a unified dual-projection model, in which each 3D point is simultaneously constrained at both row-dependent and reference timestamps along a shared continuous trajectory, enforcing stronger geometric and temporal consistency. Extensive simulations analyse the applicability of several implementations under varying conditions, and real data experiments demonstrate the accuracy, robustness, and practical effectiveness of the proposed approach.
Authors: Mohammad Amanour Rahman
Abstract: Continual learning in clinical imaging faces a dual challenge: a model must assimilate knowledge from new anatomical domains while retaining representations learned from prior tasks, a problem known as catastrophic forgetting. Existing mitigation strategies, including regularization and knowledge distillation, treat all spatial regions equally, ignoring the fact that prediction uncertainty is strongly correlated with the propensity for forgetting. We introduce UCBound-Net, a continual segmentation framework that exploits Monte Carlo (MC) Dropout uncertainty as a spatial proxy for forgetting risk. Our method contributes three synergistic components: (i) uncertainty-weighted boundary distillation, which amplifies the knowledge transfer signal at high-entropy regions of the frozen teacher; (ii) uncertainty-calibration regularization, which explicitly penalizes overconfident erroneous predictions; and (iii) uncertainty-guided exemplar selection, a memory buffer that preferentially stores samples whose boundary regions exhibit the highest predictive entropy. Evaluated on a sequential domain-incremental benchmark comprising breast ultrasound (BUSI, Task 1) followed by thyroid ultrasound (TN3K, Task 2), UCBound-Net reduces forgetting relative to naive fine-tuning, achieving a backward transfer (BWT) of -0.098 compared with -0.173, while obtaining an average Dice Similarity Coefficient (DSC) of 0.755 across both tasks. The proposed framework outperforms baseline methods without requiring task-boundary supervision. An ablation study further demonstrates that each component contributes independently to forgetting mitigation, providing a practical pathway toward uncertainty-aware continual learning for clinical image segmentation.
Authors: Mohamed Farag, Genc Hoxha, Yahia Maleki, Chris McCool, Ribana Roscher
Abstract: Reliable decision-support in digital agriculture requires accurate predictions and well-calibrated uncertainty estimates, particularly for dense prediction tasks such as semantic segmentation. Ensemble methods provide strong uncertainty quantification, but their computational and memory demands limit practical use, while single-model approximations often trade off uncertainty quality for efficiency. We propose ST-LoRA, a parameter-efficient ensemble framework that builds diverse ensemble members from a single training trajectory by combining Low-Rank Adaptation (LoRA) with snapshot ensembling. Each member shares a frozen pretrained backbone and differs only in lightweight low-rank adapters, reducing trainable parameters to under 10% of the full model while preserving ensemble diversity. We evaluate across two agricultural datasets - GrowliFlower-L (cauliflower, open field) and BUP20 (sweet pepper, glasshouse) - using SegFormer and Mask2Former, covering in-distribution performance, calibration under distribution shift, and out-of-distribution detection. Ablations show feed-forward layers, not attention layers, are the critical LoRA target for dense prediction, contrary to the attention-only convention from language models. ST-LoRA matches or exceeds full-rank ensembles in segmentation accuracy and calibration across both datasets and architectures, while substantially reducing training time, inference latency, memory footprint, and storage requirements. Against efficient baselines - Snapshot Ensemble, MC Dropout, and Deep Deterministic Uncertainty - ST-LoRA consistently matches or outperforms them in image/pixel-level OoD detection, calibration stability under shift, and cross-seed variance, with far fewer parameters and lower compute. These results show LoRA-efficient ensemble adaptation is a highly effective, practical approach for uncertainty-aware agricultural vision systems.
Authors: Omid Nejati Manzari, Guillaume Lajoie, Hassan Rivaz
Abstract: Hard symbolic-reasoning tasks such as Sudoku, maze pathfinding, and ARC remain challenging for LLMs due to their fixed-depth autoregressive reasoning, which limits systematic search, refinement, and backtracking. While recursive models such as Hierarchical Reasoning Model (HRM) and Tiny Recursive Model (TRM) address this limitation through iterative latent-state refinement, they are typically task-specific and do not leverage pretrained language priors. We propose R-Qwen, a recursive reasoning framework built upon a pretrained Qwen backbone. R-Qwen repeatedly refines a candidate solution through programmatic self-recursion and deep supervision, combining the structured iterative computation of recursive models with the linguistic and reasoning priors of pretrained LLMs. We further adapt Hierarchical Supervision Weighting (HSW) to autoregressive models by exponentially weighting losses across recursive steps. HSW reduces gradient variance by at least 50\%, improves the signal-to-noise ratio of stochastic gradients, and accelerates convergence. Across eight challenging benchmarks, R-Qwen consistently outperforms prior recursive reasoning models and substantially larger LLMs while using a comparable number of trainable parameters. Notably, on ARC-AGI dataset, our model achieves a 27.6\% improvement over the baseline, highlighting the effectiveness of recursive refinement for general symbolic reasoning. These results suggest that recursive reasoning mechanisms and pretrained language model priors are complementary approaches for improving symbolic puzzle-solving. Code and models will be released after acceptance.
Authors: Pou-Chun Kung, Aryaman Rao, Utkrisht Sahai, Hemanth Murali, Yi Liu, Rui-Yu Lin, Katherine A. Skinner
Abstract: Vision-language models (VLMs) are emerging as a key component of embodied intelligence, with growing applications in auto-labeling and end-to-end autonomous driving. However, existing approaches for improving spatiotemporal reasoning in VLMs often rely on complex preprocessing pipelines, expensive human annotations, or synthetic data, which limit scalability and introduce potential sim-to-real gaps. Moreover, although these methods have improved spatiotemporal understanding, they still lack strong metric reasoning capabilities for dynamic scenes, such as estimating object motion in real-world units. Prior work has explored LiDAR-based metric depth supervision to enhance spatial perception, but it does not directly address temporal reasoning. We introduce STAR-VLM, an automotive radar-supervised framework that enhances spatiotemporal VLMs with motion reasoning and metric velocity estimation for autonomous driving. Automotive radar is a low-cost and widely deployed sensor that provides complementary spatiotemporal supervision through range and Doppler measurements. By leveraging these measurements as label-free ground truth during training, STAR-VLM improves the metric spatiotemporal reasoning ability of VLMs. Through experiments on driving scenarios, we show that STAR-VLM achieves state-of-the-art performance on both motion classification and metric velocity estimation, outperforming even task-specific methods designed for each task. These results highlight automotive radar as a scalable and cost-effective source of supervision for building metric-aware spatiotemporal VLMs for real-world autonomous driving.
Authors: Mohamad Mofeed Chaar, Galia Weidl
Abstract: Autonomous driving (AD) systems have advanced rapidly over the past decade; however, robust perception under adverse weather conditions remains a major challenge, particularly in dense fog. In this work, we investigate fog-aware perception using synthetically generated fog data derived from the Waymo dataset. To support fog simulation, depth images are generated using an iterative learning approach. We consider five fog-density levels: clear, light fog, moderate fog, heavy fog, and very heavy fog. Instead of training a single unified model across all conditions, we train separate perception models for each fog-density level. Experimental results show that density-specific training improves performance in severe fog conditions. In particular, for the very heavy fog class, recall improves from 0.076 to 0.232, corresponding to an absolute gain of 15.6 percentage points. These findings suggest that deploying multiple specialized models, rather than a single general-purpose model, can improve perception robustness for autonomous vehicles under challenging visibility conditions. Future work will extend this strategy to additional sensing modalities, including LiDAR and radar, and evaluate generalization across diverse weather scenarios.
Authors: Jijian Zhao
Abstract: Dynamic 4D Gaussian Splatting has emerged as an efficient representation for dynamic novel view synthesis through explicit scene modeling and real-time rendering. However, existing methods typically require dense multi-view videos for sufficient geometric constraints, making capture expensive and limiting sparse-camera deployment. Reducing input views lowers acquisition cost but weakens geometry supervision, often causing missing structures and floating Gaussians. Depth priors provide geometric cues, yet no single source offers both dense coverage and reliable geometry. Monocular depth provides dense structure but is scale-ambiguous and locally biased, whereas multi-view geometric depth provides incomplete anchors consistent with the reconstruction coordinate system. To exploit their complementarity, we propose D$^2$-4DGS, a sparse-camera dynamic 4D Gaussian Splatting framework guided by dual-source depth priors. We align monocular estimates with valid multi-view geometric depths and verify their consistency to identify reliable geometric anchors. These verified anchors support consistency-aware pruning and depth supervision, while verified geometric depths and aligned mono-only estimates provide candidate geometry for densification in under-reconstructed regions. Finally, RGB-D joint optimization improves appearance fidelity and geometric consistency under sparse-view supervision. Across all nine dataset--view settings, D$^2$-4DGS achieves the highest PSNR, improving by 1.33 dB on average over the best competing method in each setting.
Authors: Dang P. M. Cao, Hieu Pham
Abstract: Cardiac digital twins convert clinical images into physiological measurements through observation operators, yet calibration studies often assume a fixed reference convention. Across four shared-backbone echocardiographic EF front-ends, phase conditioning appears to remove CAMUS baseline bias. Matched-reference analysis rejects this gain: singleplane ground-truth EF error is statistically indistinguishable across models, while single-plane ground-truth EF exceeds CAMUS biplane clinical EF by +6.30 points, explaining nearly all baseline bias. A prespecified EchoNet-Dynamic replication, with released data and our extractor aligned to the apical four-chamber plane, removes baseline overestimation and reverses the CAMUS ranking. We also quantify haemodynamic effects, conformal residual-width budgets, and EF-stratum changes, yielding a Convention-Aware EF Audit protocol that separates genuine observation operator calibration from measurement artefacts. GitHub: EjectionFraction-Bias-in-Cardiac-Digital-Twin.git
Authors: Ashfak Yeafi, Mehedi Hasan, Md Khairul Islam
Abstract: Vision-Language Models (VLMs) face a critical computational bottleneck when processing high-resolution imagery due to the $O(N^2)$ memory complexity of Softmax Multi-Head Attention (MHA). While substituting MHA with independent Multi-Layer Perceptrons (MLPs) achieves $O(N)$ scaling, it strips the architecture of spatial sequence routing, severely degrading global scene understanding and object permanence. In this paper, we propose the Linear Multi-Timescale Retention (LIA-MTR) module, a memory-efficient cross-modal bridge. By integrating an ELU-based positive feature mapping with adaptive write-gating and log-linearly distributed recurrent decays, LIA-MTR mathematically compresses continuous visual sequences into bounded memory states. Theoretical analysis proves the architecture operates with strict $O(N)$ sequence-interaction complexity. Empirically, synthetic retrieval evaluations demonstrate that LIA-MTR flawlessly routes context across 16,000 tokens, eliminating the "Lost in the Middle" degradation typical of naive linear attention. Hardware benchmarking reveals infinite-context scaling capabilities, natively processing 262,144 visual patches within an 11.2 GB VRAM footprint, whereas standard MHA suffers out-of-memory failure at 16,384 patches. Furthermore, following instruction tuning on 665K conversational samples, LIA-MTR significantly outperforms an industry-standard MLP baseline on the MME benchmark (71.00% vs. 68.11%), driven by a 10% absolute improvement in object permanence and superior global semantic extraction. This work establishes a mathematically rigorous, computationally flat foundation for infinite-context Vision-Language integration.
Authors: Zhixue Fang, Zhimin Zhang, Bi'an Du, Zijie Meng, Yan Zhou, Wei Hu, Guoxin Zhang, Pengfei Wan, Kun Gai
Abstract: Video motion transfer aims to animate a target object using dynamics from a reference video. Existing formulations largely rely on fixed structural correspondence, which becomes ill-defined when reference and target objects differ substantially in morphology, articulation, or deformation mechanisms. We introduce Motion Beyond Morphology, a perspective that seeks to transfer motion beyond fixed structural correspondence, by preserving dynamics that remain meaningful across different target morphologies. To realize this, we propose a two-stage framework. Stage~I learns complementary multi-granularity abstract motion views and uses them to bootstrap cross-category video pairs that preserve transferable dynamics across diverse morphologies. Stage~II internalizes this supervision into direct reference-video-conditioned generation, removing the need for explicit motion extraction at inference. We further introduce OpenVMT-Dataset and OpenVMT-Bench for training and evaluating image- and text-conditioned motion transfer across Same, Near, and Far category gaps. Extensive experiments demonstrate state-of-the-art motion fidelity and target preservation. Project page: https://miniz233.github.io/MotionBeyondMorphology/
Authors: Qianlong Yang, Bowen Ye, Xianda Guo, Yanlun Peng, Wenke Huang, Hongyuan Zhang, Yulei Jia
Abstract: Despite the progress of multimodal large language models (MLLMs), they continue to exhibit deficiencies in visual perception. Following visual instruction tuning, internal MLLM representations rapidly deviate from their original semantic states during inference, causing severe information degradation. While existing methods attempt to leverage external vision foundation models (VFMs) to align internal representations, we find that direct alignment with VFMs enhances visual semantics but fails to mitigate representation deviation. To address this, we propose Spatial-Spectral Visual Anchor Learning (SSVAL). The core of SSVAL is Visual Anchor Prompt Injection (VAPI), which introduces prompts that absorb rich knowledge from external VFMs during training, enabling them to serve as stable visual anchors that mitigate representation deviation during inference. Additionally, we incorporate auxiliary spatial and frequency-domain representation alignment losses to provide complementary vision-specific supervision at intermediate LLM layers. Extensive experiments demonstrate that SSVAL significantly outperforms existing methods. Code are available on our \href{https://msls38.github.io/SSVAL/}{project page}.
Authors: Huixin Sun, Wangbo Zhao, Fanyue Wei, Qiuxia Lin, Pengzhan Sun, Angela Yao
Abstract: Egocentric visual grounding requires high-resolution inputs to localize small objects. However, scaling Multimodal Large Language Models to this domain is constrained by the excessive cost of visual token processing. We identify that current efficient strategies based on token reduction are unreliable for selecting object-centric spatial evidence. To overcome this, we propose SmartRes, a framework that performs efficiency optimization in the pixel space via dynamic resolution routing. SmartRes first encodes a low-resolution view for global context and uses a lightweight router to activate high-resolution patches in object-centric regions and constructs an order-preserving visual sequence. To further enable robust routing under severe foreground-background imbalance, we introduce a margin-regularized routing objective that increases foreground-background logit separation and improves foreground recall. Experiments on Ego4D and EgoIntention show that SmartRes reduces visual tokens by up to 67% while retaining 86.4% of full-resolution performance, and achieves up to 1.66X faster inference than state-of-the-art token reduction methods with higher accuracy. Furthermore, strong performance on small object grounding indicates the effectiveness of SmartRes towards egocentric applications. Code will be publicly available.
Authors: Xiangyue Zhang, Jianfang Li, Jiaxu Zhang, Kaixing Yang, Steven Hoi
Abstract: Real-time co-speech gesture generation must produce 3D motion clip by clip as speech arrives. Existing streaming methods are open-loop: each clip depends on past context, but the model cannot check or correct its trajectory. Small errors therefore accumulate and cause drift over long sequences. We observe that this failure is mainly caused by the lack of a forward constraint rather than poor short-clip quality. A plausible key pose at the end of each clip provides a destination anchor that limits drift. Based on this observation, we propose StreamTalk, a closed-loop framework with a periodic generate-retrieve-refine cycle. Streaming Pose-Guided Generation first predicts a coarse clip, retrieves a plausible tail pose from a speaker-specific motion database, and refines the clip using this pose before continuing to the next window. During training, Stochastic Anchor Masking randomly masks pose and translation frames, teaching the model to recover complete motion from sparse boundary conditions. A part-aware DiT separates hand, body, and translation streams to reduce interference between global displacement and local articulation. On BEAT2, StreamTalk achieves state-of-the-art FGD, reduces long-horizon drift relative to open-loop baselines, and runs in real time at 76 FPS. Project page: https://xiangyue-zhang.github.io/StreamTalk/.
Authors: Yu Chen, Xiaohong Li, Xiaole Wang, Jianjin Zhang, Jun Sun, Yafeng Deng
Abstract: In video understanding, vision-language models (VLMs) must ingest massive numbers of visual tokens, causing the computational and memory cost of the prefill stage to rise sharply. Such visual sequences are highly redundant along the spatio-temporal dimension, yet a high compression ratio is often accompanied by the loss of critical details. Existing token-compression methods either employ heuristic, training-free compression with limited content adaptivity or introduce additional modules that require expensive alignment training, leaving the trade-off between efficiency and adaptivity unresolved. To alleviate this limitation, we propose CRAFT: Compression via Recursive Adaptive Fusion of Video Tokens. CRAFT recursively merges tokens by decoupling parameter-free token selection from learnable token fusion: global similarity determines which tokens to merge, while a position-aware weighting module and a content-adaptive channel-wise gate learn how to fuse them. The whole compression pipeline is query-agnostic. Because every retained token is a linear combination of the original tokens, CRAFT preserves their true spatio-temporal coordinates and stays aligned with the pre-trained language model's input distribution. Experiments on multiple representative video benchmarks show that CRAFT consistently outperforms prior state-of-the-art token-compression methods. At about $8\times$ compression, it retains roughly $97\%$ of the backbone's average accuracy and shows significant efficiency improvement.
Authors: Changhao Song, Yuxuan Wang, Qibiao Li, Youcheng Cai, Ligang Liu
Abstract: Feed-forward 3D Gaussian Splatting enables efficient novel-view synthesis without per-scene optimization, but most existing methods assume a fixed set of context views and process them jointly. This limits their applicability to online scenarios where calibrated views arrive sequentially and the scene must be updated causally. We present \emph{StreamSplat}, a streaming feed-forward 3DGS framework that incrementally maintains a persistent geometry-grounded scene state and decodes it into renderable 3D Gaussians after each input chunk. StreamSplat centers on a \textbf{Voxel-Aligned Causal Cache (VACC)}, which stores historical 3D tokens in a memory-bounded voxel structure so that memory grows with explored scene geometry rather than stream length. To better reuse history during causal prediction, we introduce \textbf{History-Projected Depth Anchoring (HPDA)} to project cached geometry as depth guidance for current cost-volume estimation, and \textbf{Cache-Guided Feature Injection (CGFI)} to inject cached latent evidence into Gaussian-token regression. Experiments on DL3DV, RealEstate10K, and ScanNet show that StreamSplat remains competitive with state-of-the-art feed-forward 3DGS methods under sparse causal inputs, despite not using future views or full-scene context. More importantly, it scales to long input streams with 256, 512, and 1024 views where fixed-view baselines run out of memory, yielding sustained improvements in novel-view synthesis quality as more observations arrive. The code will be made publicly available upon acceptance.
Authors: Fan Wei, Siru Zhong, Runmin Dong, Miao Yang, Zhaoyang Luo, Haohuan Fu
Abstract: Long-video question answering requires identifying sparse yet critical evidence from videos containing thousands of frames under a constrained visual-token budget. Existing methods either select query-aware frames in a single pass or rely on timestamped text solely as retrieval guidance, leading to two key limitations. First, selected frames tend to cluster around local relevance peaks, and once the budget is exhausted, omitted evidence cannot be recovered. Second, textual and visual evidence remain weakly aligned. We propose GCR, a training-free framework that casts fixed-budget frame selection as a joint evidence curation problem. Ground converts timestamped text into temporal events, selects query-relevant real frame anchors, and renders each event text onto its temporally aligned frame. Cover supplements grounded events with direct visual anchors for complementary visual evidence and applies global maximal marginal relevance to preserve diverse context. Refine revisits omitted temporal regions and replaces the weakest revisable context frame with a real-frame medoid---but only when the medoid offers greater evidence value. GCR maintains a fixed number of chronologically ordered frames and requires no VLM training or architectural modification. Experiments on LongVideoBench and Video-MME, across three 7B backbones and frame budgets of 8, 32, and 64, demonstrate consistent improvements in long-video QA. With the 7B LLaVA-OV backbone and 32 frames, GCR achieves 64.25% and 62.15% on the two benchmarks, outperforming the strongest reproduced baselines by 2.54 and 1.93 percentage points, respectively.
Authors: Yaning Zhang, Jiao Wu, Zan Gao, Linlin Shen
Abstract: The challenge of fair deepfake detection (FDD) has attracted increasing attention. Existing fairness-enhanced detectors often suffer from suboptimal generalization to unseen manipulations and fairness across demographic groups. They are typically developed and evaluated on demographically imbalanced distributions, resulting in biased predictions toward minority groups. In this paper, we construct a novel demographically balanced FDD benchmark to train and evaluate the fairness of detectors under both balanced and imbalanced population scenarios. Additionally, we introduce a novel expression and demographic perceptual vision-language model, termed FairForensics, for generalizable fair deepfake detection. FairForensics conducts face forgery generalization enhancement and demographic-aware fairness regularization. During face forgery generalization enhancement, built upon the novel observation of significant distribution differences between pristine and forged expression vectors, we design an expression encoder to capture high-level expression-guided forgery patterns, and an expression-perceptual visual encoder that integrates global appearance and expression forgery features while mitigating identity bias using an identity-aware patch perturbation module. Under demographic-aware fairness regularization, we propose a demographic-guided language encoder to extract population-aware global language embeddings, which boosts the decoupling of forgery features from demographic information via vision-language alignment. We devise a population-aware prototype fairness objective to enforce both inter-class separability and intra-class alignment across demographic subgroups. Extensive experiments on our balanced demographic benchmark show that our method achieves the state-of-the-art in terms of generalization and fairness.
Authors: Rahul Venkataramani, Rachana Sathish
Abstract: Promptable segmentation foundation models (FMs) such as SAM3 and Medical SAM3 promise few-shot, interactively-specified segmentation for medical imaging through a natural language interface, yet their performance on clinical tasks falls well short of this promise. We posit that this shortfall is not an artefact of insufficient medical pretraining or imperfect prompt phrasing, but a structural limitation that will persist in any domain where paired image-text supervision is scarce, as it is across most clinical modalities. We further hypothesize that the limitation is specific to natural language as a control signal: a visually grounded prompt, learned directly from the target distribution, should recover the lost performance without additional image-text data or backbone retraining. We propose Few-Shot Concept Prompt Learning (FS-CPL), which learns a continuous concept prompt embedding $\mathbf{p}^* \in \mathbb{R}^{T \times d}$ from a small support set of $K$ image--mask pairs via mask supervision, with the encoder-decoder backbone frozen. Across four public benchmarks spanning ultrasound and endoscopy (BUSI, HC18, TN3K, CVC-Clinic), FS-CPL delivers absolute Dice improvements of up to $+0.62$ over canonical text prompts and is \emph{backbone-agnostic}: it lifts both vanilla SAM3 and the domain-specifically pretrained Medical SAM3, showing that visual concept prompting is complementary to in-domain pretraining.
Authors: Mohamed Basem, Vincent Christlein
Abstract: We present our ImageCLEF 2026 Multimodal Reasoning system for the Visual Multiple Choice Question Answering (Visual MCQ) and Visual Open Question Answering (Visual OpenQA) subtasks. The challenge requires reliable reasoning over multilingual educational and scientific images with dense text, diagrams, charts, tables, formulas, and units, while enforcing strict answer formats. Our central finding is that robust output control is as important as model choice. For Visual MCQ, we replace fragile free-form generation with direct candidate label scoring from vision-language model logits, then combine complementary runs through score fusion and voting. For Visual OpenQA, we use image enhancement, concise final answer prompting, deterministic decoding, and targeted post-processing to remove reasoning traces and formatting artifacts. Without task-specific model training, our official submissions achieved third place in Visual MCQ with 0.7108 accuracy and first place in Visual OpenQA with 0.6488 COMET, 0.1391 BLEU, 0.2762 ROUGE L, and 0.2383 METEOR. The results highlight the practical value of inference engineering: careful scoring, ensembling, prompting, and cleanup can turn strong VLMs into reliable competition systems.
Authors: Rishov Paul, Frederick H. Epstein, Miaomiao Zhang
Abstract: Myocardial strain analysis of cardiac magnetic resonance (CMR) images provides an important tool for evaluating cardiac function. However, current techniques require either human-adjusted post-processing with suboptimal regional accuracy, or specialized and costly imaging acquisitions. In this paper, we propose to leverage the power of generative models to synthesize high-quality motion-derived strain values from routinely acquired CMR sequences. Specifically, we develop a novel Brownian bridge diffusion model in motion space to learn the probabilistic mapping between standard CMR motion estimated from widely adopted registration methods and highly accurate motion provided by advanced strain imaging techniques. To promote the fidelity of anatomical structure in the generation process, our model is conditioned on the corresponding CMR images. We validate our method on large-scale multi-center CMR datasets including subjects of paired standard cine CMR and advanced strain imaging acquisitions. Experimental results demonstrate that our framework significantly improves the accuracy of motion prediction and strain analysis from standard CMRs compared to existing learning-based approaches. Our research represents a new paradigm for potentially developing cost-effective, clinically deployable AI tools for cardiac function assessment with enhanced strain accuracy in busy clinical workflows. Our code is publicly available at anonymous.4open.science/r/Brownian-Bridge-strain-analysis-1140.
Authors: Masahiro Oda
Abstract: In recent years, generative AI has attracted significant public attention, and its use has been rapidly expanding across a wide range of domains. From creative tasks such as text summarization, idea generation, and source code generation, to the streamlining of medical support tasks like diagnostic report generation and summarization, AI is now deeply involved in many areas. Today's breadth of AI applications is clearly distinct from what was seen before generative AI gained widespread recognition. Representative generative AI services include DALL-E 3 (OpenAI, California, USA) and Stable Diffusion (Stability AI, London, England, UK) for image generation, ChatGPT (OpenAI, California, USA), and Gemini (Google, California, USA) for text generation. The rise of generative AI has been influenced by advances in deep learning models and the scaling up of data, models, and computational resources based on the scaling laws. Moreover, the emergence of foundation models, which are trained on large-scale datasets and possess general-purpose knowledge applicable to various downstream tasks, is creating a new paradigm in AI development. These shifts brought about by generative AI and foundation models also profoundly impact medical image processing, fundamentally changing the framework for AI development in healthcare. This paper provides an overview of diffusion models used in image generation AI and large language models (LLMs) used in text generation AI, and introduces their applications in medical support. This paper also discusses foundation models, which are gaining attention alongside generative AI, including their construction methods and applications in the medical field. Finally, the paper explores how to develop foundation models and high-performance AI for medical support by fully utilizing national data and computational resources.
Authors: Ruifeng Wang, Di Yang, Jiangtao Wang
Abstract: Player-centric ball action spotting requires temporally precise event detection together with actor attribution in crowded, partially observed multi-agent sports videos. Existing Denoising Sequence Transduction (DST) baselines treat the player-role dimension as part of a flattened frame-level representation, which weakens the inductive bias for modeling player-specific temporal evolution and inter-player interactions. To address this limitation, we propose Multi-Entity Denoising Sequence Transduction (ME-DST). ME-DST keeps the role-slot dimension throughout encoding. It uses temporal attention to model the history of each role slot, and spatial attention to exchange information across role slots at each frame. This factorized design gives the model a direct structure for separating within-player evolution from inter-player context. We also add learnable role embeddings, tracking-derived tactical features, and fused visual predictions from X3D-L and Swin3D-S. Experiments on the FOOTPASS dataset show that ME-DST reaches a Micro F1 of 0.778. This improves the strongest official TAAD+DST baseline by 10.3 percentage points. Controlled ablations show that preserving the entity axis and encoding role identity are central to this gain. These results suggest that explicit entity modeling is an effective inductive bias for player-centric sports event understanding.
Authors: Inha Lee, Dongjae Jeong, Junhee Lee, Kyungdon Joo
Abstract: Recent geometric foundation models enable feed-forward inference for SLAM, but their predictions are strongly dependent on the input view set, which leads to geometric inconsistencies and trajectory drift when results are chained over long sequences. Online deployment further exposes a trade-off between the low latency of two-view tracking and the constraint richness of multi-view inference. We introduce UniSim-SLAM, an integrated system that runs lightweight two-view keyframe tracking in the frontend and performs periodic multi-view submap refinement in the backend. To combine predictions defined in heterogeneous local coordinates with inconsistent scales, we formulate a unified multi-level factor graph on $Sim(3)$ that jointly optimizes global keyframe poses and submap poses. The graph integrates temporal view-to-view odometry edges, view-to-submap bridge edges with depth-statistics scale anchoring, and submap-to-submap tie and scale constraints to enforce consistent similarity relations across submaps. Experiments on TUM RGB-D and 7-Scenes show that UniSim-SLAM achieves state-of-the-art accuracy in the uncalibrated setting, reducing trajectory error by $38.5\% $ on TUM RGB-D and $45.9\%$ on 7-Scenes compared to prior best results. Project page: https://vision3d-lab.github.io/unisim-slam/
Authors: Hai Nguyen, Tung Vu, Cong Tran
Abstract: Vision-language models (VLMs) achieve strong semantic understanding but remain unreliable in metric spatial reasoning, particularly when queries require comparing multiple instances of the same object category. We study this problem through the Closest-Instance Distance Query (CIDQ), where a model must identify the nearest visible candidate to a unique reference object and estimate their gravity-aligned floor-plane distance. We introduce SPATIALQUERY, a training- free framework for CIDQ reasoning from a single RGB image, together with SPATIALQUERY-1M, a benchmark containing over one million RGB-only question-answer pairs from 200 indoor scenes. SPATIALQUERY recovers instance-level metric geometry and transforms it into a canonical Bird's-Eye View through Scene Cubifying, which represents objects as uniformly sized, category-coded blocks to emphasize their relative floor- plane locations. We further propose Uncertainty-Aware Chain-of-Thought (UA-CoT) prompting, which incorporates geometry- derived per-instance uncertainty into the VLM reasoning process. Without task-specific fine-tuning or architectural modification, SPATIALQUERY with Qwen3-VL-8B achieves a Floor-MAE of 0.259 m, an Unc-Acc@0.3 m of 90.5%, and a proximity-decision accuracy of 84.18%, outperforming fine-tuned spatial specialists, general-purpose VLMs, and closed-source frontier models. Code, benchmark resources, and an interactive demo are available at https://namhai1810.github.io/SpatialQuery/.
Authors: Shanaka Ramesh Gunasekara, Akila Eranda Devanarayana, Imasha Guruge, Nuwantha Fernando, Ehsan Asadi
Abstract: Accurate crack assessment in electroluminescence (EL) images is important for photovoltaic (PV) reliability analysis, yet existing segmentation methods often fail to capture the thin, elongated, and structurally constrained nature of crack defects. This paper proposes a Solar Topology Crack Network (STC-Net) that incorporates edge priors, spectral priors, and a boundary-topology refinement module to improve crack continuity and boundary preservation. The framework further extends segmentation to power-loss estimation by deriving a crack-associated inactive-area proxy from the predicted masks. Experiments on the PVEL-S dataset show that STC-Net achieves 95.98 MIoU, 98.01 MDice, and 98.00 MAcc during training, and 72.52 MIoU and 80.16 MDice on unseen test samples. These results demonstrate that STC-Net provides accurate crack localization while offering a practical link between EL-based defect segmentation and PV degradation assessment.
Authors: Zepu Wang, Jingze Liang, Weijie Xiao, Kexin Chen
Abstract: Extremely low-light RAW enhancement aims to recover severely attenuated sensor signals, yet existing methods often focus on illumination and noise while overlooking the motion-induced degradations inherent in practical low-light imaging. We present a framework for robust extremely low-light RAW enhancement under realistic acquisition degradations. First, we introduce See in the Degraded Extremely Dark (SIDED), a new dataset that applies controlled motion degradation to extremely low-light RAW pairs while retaining their original sensor noise. Second, we propose a unified RAW tokenizer equipped with explicit domain-conditioned representation calibration to align extremely low-light and well-exposed RAW data, followed by a MeanFlow that performs enhancement in a single function evaluation. To our knowledge, this is the first work to formulate extremely low-light RAW enhancement under realistic motion-degraded acquisition and address it with MeanFlow. We further introduce a physics-guided refinement model to strengthen illumination--reflectance consistency, pixel fidelity, and color preservation without incurring additional inference cost. Extensive experiments demonstrate that our framework achieves state-of-the-art performance in extremely low-light RAW enhancement, and robustly handles coupled motion and noise degradations.
Authors: Yuxin Yao, Kendong Liu, Shiqi Zhou, Jiazhi Xia, Junhui Hou
Abstract: 3D Gaussian Splatting has achieved remarkable success in photorealistic and efficient rendering, leading to a rapid increase in 3D assets represented by 3D Gaussian primitives. Directly rigging these assets with arbitrary skeleton topologies is highly desirable. However, training a feed-forward skinning framework is infeasible due to the lack of high-quality 3D Gaussian rigging datasets. An alternative solution is to transfer mesh-based techniques to 3D Gaussian-based representation, but 3D Gaussian primitives are not restricted to the surface and lack explicit topological connectivity. Moreover, this kind of method suffers from poor generalization to unseen data due to its strong dependence on training data, while acquiring high-quality rigging data is prohibitively expensive. To address this challenging problem, we propose G-Skin, a novel generative skinning framework designed for expressive and high-fidelity animation with 3D Gaussian representation. To overcome this 3D data scarcity, we introduce a skeleton-controllable image generation model leveraging 2D vision foundation models to distill powerful motion priors into pseudo-guidance. Guided by these priors, we formulate an optimization pipeline incorporating geometry-aware regularizations, which stabilizes the learning process and ensures smooth, structurally coherent skinning weights. G-Skin also generalizes flexibly to the augmented variants of 3D Gaussian representation designed to mitigate animation-induced rendering artifacts. Extensive experiments validate the effectiveness of our approach, demonstrating clear advantages over state-of-the-art methods. Project page: https://yaoyx689.github.io/GSkin.html.
Authors: Hakan Emre Gedik, Shashank Gupta, Alan Bovik
Abstract: No-reference image quality assessment (NR IQA) has recently benefited from deep and multimodal models, yet many SOTA systems still violate at least one basic requirement: they either discard critical quality cues via aggressive resizing, fail to generalize across resolutions, cannot be jointly trained on heterogeneous IQA datasets with mismatched MOS scales, or require prohibitive computation. We present \textbf{ReLIQS}, a model for \textbf{Re}solution-agnostic \textbf{L}earning for \textbf{I}mage \textbf{Q}uality with \textbf{S}aliency, which is resolution-agnostic, preserves original-resolution quality cues, learns from multiple subjective studies, and remains computationally efficient and budget-adaptive. ReLIQS is a CLIP-based multiscale patch-driven architecture that learns both \emph{where to look} and \emph{how to judge} quality. Fixed-size patches are sampled across multiple resolutions, including the original resolution, and encoded with a CLIP vision backbone. A lightweight Perceptual Importance Estimator then predicts IQA-specific importance maps to select a small set of informative patches, and a Latent Quality Axis Module aggregates their embeddings into a single image-level score. Across authentic, synthetic, and AIGC benchmarks spanning diverse resolutions and distortions, ReLIQS generalizes better than strong CNN-, CLIP-, and MLLM-based baselines with matching or reduced computational cost.
Authors: Nayoung Kim, Nan Jiang, Bangjie Sun, Jaewon Shin, Sojeong Kim, Jun Han
Abstract: As digital drawings are increasingly shared online, reliable authorship verification has become important for protecting artists and resolving disputes. Yet when authorship is questioned, verification may have to rely only on the disputed drawing and reference drawings known to be created by the claimed artist. This setting is challenging for two reasons. First, artist-specific drawing behavior, such as pen pressure and movement speed, is informative but is not available from a completed drawing. Second, similarities in the depicted object or scene can obscure similarities arising from the artist. We propose IDraw, a framework that learns from drawings paired with tablet-pen sensor signals collected from separate training artists. This allows IDraw to infer drawing behavior from completed images during a later authorship dispute, without requiring sensor data from the artist being verified. IDraw also reduces the influence of drawing content by identifying information shared by drawings of the same object across different artists and suppressing it before comparing drawings. To support this approach, we construct the first multimodal dataset for digital drawing authorship verification, containing 1,110 drawings from 37 artists and 14 types of tablet-pen sensor signals. Evaluated on previously unseen artists across nine image-encoder backbones, IDraw consistently outperforms standard image-based verification and reduces verification error by up to 40%. These results demonstrate that inferring drawing behavior from completed images and suppressing drawing content improve digital drawing authorship verification.
Authors: Xingyan Li, Jordan A. Caraballo-Vega, Jie Gong, Mark L. Carroll, Jianwu Wang
Abstract: Geospatial foundation models (GeoFMs), pretrained on large-scale geospatial data such as Earth observation (EO), climate, and weather data, have shown promising performance when fine-tuned on diverse downstream tasks. However, there are two challenges of adapting EO-pretrained GeoFMs to practical downstream datasets. The first challenge is how to handle spectral mismatch: pretrained patch embeddings expect a fixed set of input bands, whereas downstream sensors may provide different channels. The second challenge is how to reduce fine-tuning cost and make it efficient. While existing work has made efforts on these challenges individually, jointly improving fine-tuning performance under spectral mismatch while reducing adaptation cost remains underexplored. We propose SPECTRA, a parameter-efficient fine-tuning framework that addresses both spectral mismatch and adaptation cost. To handle spectral mismatch, SPECTRA introduces Band-Routed Embedding (BRE), which maps all available downstream bands into the band space expected by the pretrained GeoFM. By using BRE, all available bands in the downstream dataset are utilized to improve the selected-band input without changing the pretrained patch embedding interface. To reduce adaptation cost, SPECTRA further introduces a Stage-wise Transferability-aware LoRA (ST-LoRA) fine-tuning. ST-LoRA estimates stage-wise transferability before fine-tuning and assigns stage-specific LoRA ranks, concentrating trainable parameters on the stages with high transferability for the target task. Across three EO-pretrained GeoFMs and four downstream segmentation datasets, experiments show that BRE improves performance by utilizing all spectral bands, while ST-LoRA reduces trainable parameters compared with full fine-tuning and standard LoRA. Code is available at https://github.com/big-data-lab-umbc/SPECTRA.
Authors: Xiaohao Yang, Aohua Tian, Derek Van Berkel, Xu Qiang, Mark Lindquist
Abstract: Addressing urban blight has seen increased focus in the past 15 years. Assessing urban blight is essential for guiding urban planning, targeting rehabilitation, and safeguarding public health, yet traditional residential blight surveys are difficult to maintain at scale due to the labor-intensive cost and long-term cycle. This study introduced a scalable framework for estimating residential blight using open-source large vision-language models on multiple views. Structured prompts guided models to evaluate housing attributes, including roof integrity, wall damage, and broken or boarded openings, producing both binary assessments and probabilistic estimates of disrepair. To evaluate the performance of these visual assessments, we compared professional human annotations of these features across several models, including an ensemble stacking approach based on XGBoost and a weighted scoring system. Results showed that (i) multiple street views can contribute to the improvement of accuracy, (ii) large vision-language models have different strengths of inference, (iii) the ensemble learner outperforms individual base models, enhancing robustness across all residential conditions and blight assessment. The practical application of the method allows low-cost tracking and management of housing stock conditions, providing a regularly updatable complement to traditional blight surveys.
Authors: Dayu Li, Shihao Zhou, Leizhi Shu, Jin Wu, Chi Man Vong, Jufeng Yang
Abstract: Adverse weather image restoration aims to recover clear visibility from degraded images in complex weather conditions. Existing works attempt to address this problem by modeling relationships between pixels, however, this paradigm defies the spatially non-uniformity fact of degradations and learns non-discriminative features from semantic-conflict regions. In this paper, we propose SSR, a \textbf{S}emantic-center guilded \textbf{S}tate space model for image \textbf{R}estoration. The key idea of SSR is to shift the conventional scanning strategy of pixel-serial to semantic-guilded one. Specifically, we introduce a Superpixel-guided Selective Scan Mechanism ($\text{S}^3$M), which first partitions the image into perceptually coherent regions via superpixel clustering and then performs relations modeling within the semantic-related regions. Moreover, a Region-level Gating Mechanism (RGM) is developed to perform intra-region calibration by modulating degradation outliers within each semantic superpixel unit along the channel dimension. Extensive experiments on \textbf{6} well-established benchmarks demonstrate that SSR performs favorably against state-of-the-art models with competitive computational cost.
Authors: Siying Li, Ying Ni, Jie Sun, Jian Sun, Haotian Shi
Abstract: End-to-end (E2E) autonomous driving algorithms require rigorous closed-loop validation in simulation environments offering high visual fidelity, strong interactivity, and real-time performance. Existing approaches, from game engines to static neural rendering, inherently trade off these requirements and struggle with the dynamic scene composition essential for E2E testing. To bridge this gap, we propose a novel decoupled 3D Gaussian Splatting (3DGS) framework tailored for large-scale E2E evaluation. We fundamentally decompose scenes into a high-fidelity static background and manipulable dynamic agents using an object-centric canonical representation. To resolve resulting representational conflicts, we introduce three targeted modules: (1) asset compression via perceptual pruning and vector quantization for real-time traffic rendering; (2) map-guided geometric registration leveraging semantic topology to strictly align trajectories; and (3) proxy-based relighting transferring ambient illumination for seamless photometric integration. Extensive experiments demonstrate that DecoupleGS achieves a balanced fidelity-efficiency trade-off, improves metric and photometric consistency, and provides a practical closed-loop sensor simulation platform for E2E autonomous driving evaluation.
Authors: Yue Ma, Jiangming Wang, Yucheng Wang, Xilai Wang, Zhiyuan Li, Xinyu Wang, Hongyu Liu, Ruofan Liang, Songchun Zhang, Yuxuan Xue, Qifeng Chen
Abstract: We present LiveLight, the first diffusion-based framework for real-time streaming video relighting with interactive 3D lighting control. Achieving this is non-trivial, as it requires overcoming three critical challenges: effectively injecting dynamic 3D lighting into a diffusion model, maintaining high-fidelity generation under an extremely low NFE (Number of Function Evaluations) budget for real-time speed, and facilitating continuous streaming for interactive control. To address these pain points, we propose three key designs. First, for accurate lighting injection, we propose a lightweight adapter that feeds Multi-Plane Light Irradiance (MPLI) conditions-depth-aware irradiance maps encoding 3D lighting geometry-directly into the diffusion backbone. Second, to prevent rendering quality degradation at low NFEs towards real-time distillation, we introduce a geometry-guided feedback branch. This training-time constraint leverages a frozen geometry estimator to enforce depth- and normal-consistent relighting, ensuring geometrically plausible shading without adding inference overhead. Finally, to enable streaming interaction, we develop a progressive rolling-window strategy that maintains a denoising ladder of latent chunks at varying noise levels. By propagating intermediate states, this strategy guarantees temporal coherence and supports arbitrarily long video relighting with per-frame reference refresh. Extensive experiments on real-world and synthetic benchmarks demonstrate that LiveLight achieves state-of-the-art relighting quality while running at real-time speed, significantly outperforming offline baselines in temporal stability, lighting controllability, and user preference. To foster real-time interactive relighting research, we will publicly release our models, training data, and synthetic data generator.
Authors: Junyeong Park, Sowon Min, Euna Jang, Soobin Kim, Jiho Jin, Hyunseung Lim, Gahyeon Bae, Hwajung Hong
Abstract: Text-to-image (T2I) generation models are increasingly embedded in applications such as media content creation and education, raising concerns about how their outputs may reproduce social biases. Prior work has shown that T2I models exhibit social biases, yet existing evaluations largely focus on a photo generation task. As a result, it remains unclear whether and how such biases manifest in more narrative visual formats, such as storyboards and comics, where characters and events are presented across multiple panels. In this work, we compare bias expression across photo, storyboard, and comic generation in six T2I models by adapting BBG, a text-based bias evaluation framework, to image generation. Our results show that proprietary models generate 25.9% biased outputs in photo generation on average, with biased outputs increasing by 9.6pp in storyboard generation and 18.2pp in comic generation. We also find that photos mainly encode biases through subtle visual cues, while storyboards and comics reveal them more explicitly through event sequencing, character positioning, narrative resolution, and textual elements. These findings show that biases that remain less visible in photo generation may surface in narrative visual formats, highlighting the importance of evaluating T2I systems with diverse visual formats beyond photo generation.
Authors: Jiawei Cao, Junyi Feng, Jiashen Hua, Ziheng Huang, Bing Deng, Kaijie Wu, Chaochen Gu, Jieping Ye
Abstract: Universal Multimodal Embeddings (UMEs) aim to unify various modalities and tasks into a shared representation space. In recent years, this field has witnessed substantial progress driven by the development of Multimodal Large Language Models (MLLMs). However, a crucial capability, visual identity discrimination, remains underexplored in existing UME methods, despite its critical role in a wide range of tasks, including instance retrieval, re-identification, and identity preservation in AI-generated content. To bridge this gap, we propose a unified formulation for visual identity discrimination~(VisID) and introduce $\textbf{MVEB}$ ($\textbf{M}$ultimodal $\textbf{V}$isual Identity $\textbf{E}$mbedding $\textbf{B}$enchmark), a large-scale benchmark curated from both real-world and synthetic datasets to support evaluation and training. Furthermore, we present a simple yet effective learning framework that jointly optimizes general multimodal and visual identity representations through a carefully designed identity-aware sampling mechanism. Extensive experiments demonstrate that our approach successfully endows UMEs with strong identity discrimination capability and maintains competitive general multimodal performance. We believe this work not only illuminates a critical yet neglected capability, but also takes a step toward more holistic universal multimodal embeddings. Code and data are available at \href{https://chrisclear3.github.io/MVEB}{MVEB}.
Authors: Zhaoding Ding, Chenglong Li, Jiandong Jin, Kewei Ying, Wentao Wu
Abstract: Existing RGBT trackers typically employ fusion functions with fixed parameters across different targets and scenarios. Although dynamic-architecture methods improve fusion flexibility by selecting among predefined operations, they still cannot adapt the fusion parameters to the evolving target state. To address these issues, we propose a Parameter-Dynamic Adaptive Fusion and Calibration Network (PAFCNet) for RGBT tracking. PAFCNet dynamically generates target-conditioned parameters for multimodal fusion and temporal calibration, enabling the tracking process to adapt to target appearance variations and modality quality fluctuations. Specifically, we introduce a Target-Adaptive Hypernetwork (TA-HyperNet) that leverages template representations, which preserve stable target identity and recent appearance changes with less background interference, to generate target-conditioned parameters for subsequent fusion and calibration. Based on TA-HyperNet, we design a target-aware parameter-dynamic fusion module that uses the generated parameters to modulate the fusion process. This enables the fusion module to adapt to changes in target appearance and complex scene conditions. Furthermore, since spatio-temporal information propagation may accumulate tracking noise, we propose a dynamic spatio-temporal calibration module that employs TA-HyperNet to generate calibration parameters for spatio-temporal tokens. By dynamically calibrating historical information before propagation, the module improves the reliability of temporal representations. Experimental results demonstrate that PAFCNet achieves competitive performance on multiple RGBT tracking benchmarks.
Authors: Siam Tahsin Bhuiyan, Rashedur Rahman, Sefatul Wasi, Riyadul Islam, Syoji Kobashi, Ashraful Islam, Saadia Binte Alam
Abstract: Deep learning models for medical image analysis typically apply a fixed amount of computation to every input, regardless of case difficulty. Anatomy-guided dual-stream architectures have been shown to improve diagnostic performance, but they evaluate both streams unconditionally, even on cases a single stream could already resolve confidently. We propose SecondOpinion, a framework in which a fast primary stream processes every case, while a second, anatomy-guided stream is invoked only when GateKeeper, a gating mechanism trained explicitly as a binary correctness classifier, judges that the primary stream's prediction needs additional scrutiny, much as a clinician might seek a second opinion on a difficult case. When activated, the two streams are combined through a lightweight cross-attention fusion module. We evaluate SecondOpinion on a unified five-class chest X-ray dataset and a pelvic fracture dataset, the latter including a held-out, harder subset of fractures that are invisible on X-ray but confirmed via CT. SecondOpinion matches or exceeds prior state-of-the-art performance on both tasks, while activating its anatomy-guided stream on only 9.23% of chest X-ray cases, rising to 24.12% on visible fractures and 45.71% on invisible fractures, an activation rate that tracks task difficulty directly. These results suggest that supervising a gating signal toward correctness, rather than relying on unsupervised confidence, allows a model to allocate anatomical reasoning where it is actually needed.
Authors: Yongkang Zhou, Xiang Xia, Cheng Yan, Fan Xu, Wuyang Zhang
Abstract: Diffusion vision-language models (dVLMs) iteratively denoise masked responses while conditioning each denoising step on visual evidence, making visual conditioning a substantial recurring inference cost. Unlike autoregressive decoding, diffusion generation repeatedly revisits the entire response as uncertainty evolves. Our analysis reveals that visual evidence demand is strongly step-dependent, motivating adaptive allocation across denoising steps. Existing inference acceleration methods operate through decoding-side strategies or visual token compression via pruning and merging, but do not explicitly treat visual evidence as a resource whose demand evolves across the diffusion process. Therefore, we present Denoising-Aware Visual Evidence Trajectory Allocation (DAVET), a training-free framework that allocates visual evidence according to the evolving generation state. Starting from a phase-conditioned evidence trajectory, the proposed allocation policy uses operation demand to set an evidence reserve whose allocation at each denoising step is modulated by trajectory risk. DAVET realizes the resulting budgets through a hierarchy of evidence views constructed from a single visual encoding, separating when and how much evidence is needed from how the evidence views are constructed. Evaluated on two representative dVLMs, LLaDA-V and LaViDa, across multiple visual-understanding benchmarks, DAVET achieves an average speedup of 1.55$\times$ with an average relative performance drop of 1.86\%, showing that denoising-aware visual evidence allocation can reduce visual conditioning cost while largely preserving generation quality.
Authors: Hongyi Fang, Jiahui Wu, Yichen Yue, Benjia Zhou, Dan Zeng
Abstract: Hallucination remains a persistent challenge in generative super-resolution (GSR), where reconstructed results may contain visually plausible yet weakly supported content, structural deviations, or unnatural textures with respect to the low-resolution (LR) input. Existing GSR methods have extensively explored the trade-off between perceptual realism and reconstruction fidelity, but the division between preserving reliable coarse-scale information and restoring more uncertain fine details is often handled implicitly within the overall restoration process. Visual autoregressive (VAR) modeling provides a natural opportunity to revisit this issue, as its coarse-to-fine next-scale prediction offers an explicit scale-wise generation interface. However, existing VAR-based SR methods still inherit the original full 1-to-$N$ autoregressive generation path, even though, for super-resolution, coarse-scale information in LR is often relatively more reliable, while long autoregressive chains may accumulate prediction errors. Motivated by these observations, we propose \textbf{K2N}, which reformulates VAR-based SR from full-path generation into a $k$-to-$N$ detail continuation process. Specifically, early coarse-scale states are established directly from LR, while only the remaining finer scales are restored autoregressively. Experimental results show that K2N remains competitive with the VARSR baseline on standard SR metrics, while exhibiting clearer advantages on hallucination-focused evaluation. These findings suggest that explicitly rethinking the generation path in a scale-wise manner can be a promising direction for improving the reliability of generative super-resolution. Our code will be released soon at https://github.com/BRL-SYSU/K2NSR.
Authors: Guangming Fu, Jin Song, Yiyun Fei, Guoqiu Li, Ruigao Yang, Jianan Jiang
Abstract: Part-level 3D generation has recently attracted increasing attention for producing structured and editable 3D assets. However, existing methods typically decompose objects according to functional semantics rather than the editable material boundaries (e.g., fabric, wood, metal) required in practical 3D applications such as interior design. Additionally, current methods often generate parts independently, causing computational costs to scale linearly with the part count. To address these limitations, we present PartMat, an efficient material-aware 3D part decomposition pipeline that represents multi-part geometry with a single global latent. Given a reference image and a single whole-object geometry, PartMat decomposes the object into parts that follow material boundaries. First, we propose PartVAE to learn such a unified representation and decode all material parts in a single forward pass, thereby decoupling inference cost from the number of parts. Second, with this representation, a diffusion model is trained for part generation and refined via reinforcement learning for accurate material assignment and overlap suppression. Finally, to recover fine-grained geometric details, we introduce a sparse-voxel flow-matching model with part attention for geometry post-processing. Extensive experiments demonstrate that PartMat significantly outperforms existing baselines in material-aware decomposition accuracy and achieves comparable geometric quality, while maintaining efficient inference.
Authors: Huanyao Zhang, Jiepeng Zhou, Runhao Zhao, Yanzhe Shan, Jiaoyang Chen, Bowen Zhou, Bo Li, Fang Wang, Jialong Wu, Zhengwei Tao, Lang Mei, Xiaohan Yu, Liyan Liu, Chong Chen, Wentao Zhang
Abstract: Multimodal large language models (MLLMs) have advanced visual understanding and reasoning, yet their static parametric knowledge limits their ability to address knowledge-intensive and dynamically evolving open-world problems. To move beyond this limitation, multimodal deep search has emerged as a key direction for open-world information access, evolving from single-turn factual retrieval toward long-horizon, multi-turn search guided by visual evidence. However, existing methods typically confine vision to the input or answer stage, overlooking its role in intermediate reasoning, and lack designs tailored to long-horizon interaction. Consequently, visual evidence rarely drives continued retrieval, constraining both interaction depth and reasoning span. To address these limitations, we propose DeepVoyager-VL, a long-horizon multimodal deep-search framework for vision-in-the-loop search. Specifically, we construct a multimodal event graph to drive data synthesis, yielding problems with intermediate visual dependencies and long reasoning chains. We then design an agent framework for active visual acquisition and on-demand image loading. Finally, we fine-tune models on the synthesized data without reinforcement learning. Extensive experiments across ten multimodal search benchmarks demonstrate the effectiveness of our method.
Authors: Yongcong Wang, Pu Wang, Hingchin Chen, Runci Bai, Yucheng Xin, Chen Wu, Chengchao Shen, Guangwei Gao, Siyuan Yao, Pengwen Dai, Zhuoran Zheng
Abstract: Real-world video arrives hazy, rainy, dark, or noisy, and a deployable restorer faces three demands at once: no degradation label, native 4K output, and stability in playback. Existing methods answer them separately and break on the joint problem, because per-frame degradation readings flip between frames, downsampled proxies erase the rain and noise they are meant to remove, and dense temporal alignment does not fit 4K memory. No paired benchmark even poses that problem, so we build one. UHV-4K-AIO renders physically modeled haze, rain, sensor noise, and low light over the same 100 clean 4K clips with shared depth and motion, and its construction exposes the split MoCRA is built on: haze and low light survive aggressive downsampling, while rain and noise exist only at native scale. Band-matched compositional conditioning follows, spending conditioning capacity, computation, and supervision in the band where each degradation lives. One dictionary of rank-1 atoms, recomposed sparsely per frame, conditions both a once-per-clip coarse branch and a shallow native-resolution refiner, in 3.6M parameters and with no optical flow. Trained once for all four tasks, MoCRA takes the best task-mean PSNR of eleven retrained image and video baselines, holds warping error at the level of the flow-based video models while never estimating motion, and restores native 4K in under half a second, against 1.7 seconds for the fastest baseline.
Authors: Yonghan Gao, Zehong Chen, Lijian Xu, Jingzhi Chen, Jingwei Guan, Xingyu Zeng
Abstract: Recent visual-text compression (VTC) methods, typified by DeepSeek-OCR, report impressive high token compression ratios for long-context modeling tasks by leveraging text-to-image rendering. However, existing evaluation protocols heavily rely on downstream task performance. Such evaluation metrics fail to accurately measure text preservation due to the strong inherent linguistic priors of Multimodal Large Language Models (MLLMs). In this work, we introduce a new evaluation framework that decouples MLLMs' capabilities to faithfully assess VTC quality. Within this framework, we further introduce the ZeroSense Benchmark to ensure low semantic correlation of testing samples. By eliminating textual dependencies, our benchmark guarantees that the evaluation results are purely reflective of VTC quality, unaffected by the semantic inference capabilities of downstream models. Extensive experiments across multiple datasets demonstrate that VTC quality and downstream task accuracy diverge significantly, highlighting the necessity of our decoupled evaluation framework.
Authors: Kaustubh Kapil, Kishor P. Upla
Abstract: With the widespread adoption of Vision Transformers in modern AI, the need to analyze their inherent representational behavior has become increasingly important. While most existing studies emphasize token geometries and training dynamics, the evolution of representational covariance structures and class-level geometric organization remains comparatively underexplored. In this work, we investigate semantic geometry and class separability as representations evolve across the layers of ViT-Small/16 through TGO-III: Semantic Geometry Observatory. It is a framework designed to analyze the emergence of semantic organization, feature evolution, and class-wise representation geometry throughout training. The framework employs multiple complementary observatories, including Linear Probe Accuracy, Fisher Ratio, Class Centroid Distances, Local Intrinsic Dimension, and Local PCA Rank, to quantify the progressive evolution of discriminative representations. Our analysis reveals that class representations become progressively more linearly separable, Fisher discriminability increases, class centroids move farther apart, and local representation manifolds exhibit structured class-dependent geometric complexity. These observations provide empirical evidence supporting the Semantic Expansion Hypothesis, suggesting that the manifold expansion observed in previous observatories is accompanied by the progressive organization of representations into increasingly discriminative semantic structures. Collectively, TGO-III extends the Transformer Geometry Observatory framework by establishing a direct connection between manifold geometry, covariance evolution, and semantic organization during Transformer training.
Authors: Ya-nan Guan, Shaonan Zhang, Tao Dai, Tianqu Zhuang, Yongchao Qiao, Zhensen Chen, Shu-Tao Xia, Hang Guo
Abstract: Low-light image enhancement (LLIE) seeks to restore structural fidelity, natural color rendition, and proper exposure from images captured under inadequate lighting conditions. Recent state-of-the-art approaches, such as CIDNet, adopt a dual-branch architecture comprising a chrominance (HV) branch and an intensity (I) branch to separately model decoupled chromatic and luminance information within the HVI color space. However, these methods overlook the mutual interaction between intensity and chrominance components, which inherently limits their representational capacity and leads to suboptimal enhancement performance. To address this limitation, we propose the Conditional Mutual Information-Guided Network (CMIG-Net), which leverages conditional mutual information as a principled metric to quantitatively assess the contribution of chrominance features conditioned on the available intensity information. In particular, we design a Conditional Mutual Information Calibration (CMIC) module that generates a conditional information map, enabling region-adaptive recalibration of chrominance representations according to local illumination statistics. Furthermore, we introduce a Dynamic Dual-branch Information Restoration (D2IR) module, which adaptively governs bidirectional information flow between the intensity and chrominance branches, guided by both the conditional prior and the instantaneous restoration state. Extensive experiments on paired LLIE benchmarks demonstrate that CMIG-Net consistently outperforms CIDNet, achieving up to a 0.619 dB gain in PSNR, with a 0.382 dB improvement specifically on the challenging Sony-Total-Dark dataset.
Authors: Jeonghyeok Do, Munchurl Kim
Abstract: Existing generative models for earth observation (EO) predominantly rely on fine-tuning natural image priors, which limits their scalability and introduces perspective biases that conflict with geospatial constraints. To address this, we introduce GeoCore-9B, a 9-billion-parameter generative foundation model, which is the first of its scale to be trained from scratch exclusively on EO data. Unlike previous EO foundation models, GeoCore-9B is built upon a Flow Matching-based Diffusion Transformer (DiT) and natively conditions generation on text descriptions and continuous geospatial metadata, including ground sample distances, latitudes, and longitudes. To overcome the convergence and spatial disorientation challenges of training at this scale, we propose a Geospatial Semantic Alignment loss. This objective distills structural Earth surface priors (e.g., terrain and urban areas) from a frozen specialist teacher network, constraining the diffusion latent trajectory during training without adding inference overhead. Pre-trained on the global-scale Git-10M dataset, GeoCore-9B demonstrates strong downstream versatility. Beyond standard proxy generative tasks, we show that GeoCore-9B can be effectively adapted for practical EO applications, including highly challenging tasks such as cloud removal and SAR-to-optical cross-modal translation. Extensive evaluations confirm that GeoCore-9B establishes new state-of-the-art performance in both visual fidelity and geographic structural accuracy.
Authors: Jing Wu, Jianhua Wu, Jiayi Guan, Jiahong Chen, Jinghui Lu, Hangjun Ye, Bingzhao Gao, Long Chen
Abstract: Vision-Language Models (VLMs) perform well on commonsense reasoning tasks but struggle with visual spatial reasoning. Most existing solutions introduce extra 3D prior inputs or external spatial encoders, which increase complexity and degrade the underlying VLMs' general-purpose capabilities after spatial fine-tuning. To this end, we propose a parameter-efficient \textit{\textbf{Spatio}-vision \textbf{L}anguage \textbf{M}odels (SpatioLM)}, that enhances spatial intelligence without extra 3D prior inputs or third-party spatial encoders. Concretely, we design a plug-and-play and non-invasive spatio-vision module that elicits the spatial knowledge inherent in VLMs. Furthermore, we innovatively leverage pseudo depth and camera information as supervision to guide the model in learning physically coherent representations. Extensive experiments show that SpatioLM achieves significant improvements in diverse tasks, including spatial perception and understanding while effectively limiting the degradation of general capabilities. Notably, the model achieves an impressive score of 71.6 on the VSI-Bench (the first model to surpass 70). In addition, it attains competitive performance when transferred to embodied manipulation tasks. Code is available at \href{https://github.com/xiaomi-research/spatio-lm}{\faGithub~spatio-lm}.
Authors: Zhenhao Zhang, Jiajun Zhang, Wei Min, Yebin Liu
Abstract: Hand-object interaction (HOI) is a fundamental human behavior with broad applications in AR/VR, digital humans, and embodied interaction. Existing methods typically require predefined object geometry, object trajectories, or task-specific conditions, limiting their use with natural real-world inputs. To address this, we study a more practical problem of synthesizing 3D hand-object interaction sequences from a single RGB photograph and an open-vocabulary language instruction, and introduce PhotoHOI. PhotoHOI first uses a vision-language model to parse the input image and instruction into a structured task specification, including the interaction object, target region, and spatial relation. It then recovers a compact task-relevant 3D scene and plans a smooth collision-aware object trajectory based on the recovered object states, support relations, and surrounding scene geometry. To synthesize hand motion that generalizes to real-world photographs and unseen objects, it learns transferable task-conditioned contact and contact-conditioned grasp priors from large-scale affordance and HOI data. The grasp is further refined in a learned latent space, constraining the optimization to a plausible hand-pose manifold. Experiments on GRAB and H2O demonstrate improved contact quality and reduced penetration over representative baselines. Results on real-world photographs further demonstrate higher task success and scene consistency, together with generalization to unseen objects and open-vocabulary instructions.
Authors: Huy Quang Ung, Guillaume Habault, Roberto Legaspi, Hao Niu, Lian Cao, Masato Taya
Abstract: Rapid and accurate post-disaster building damage assessment is essential, yet remains a challenging task. Unmanned Aerial Vehicle (UAV) imagery offers a timely and high-resolution view of affected areas, but existing Computer Vision (CV) models often demand large annotated datasets, generalize poorly across geographic regions and their assessment policies, and are confined to the specific tasks they were trained for. Large Vision-Language Models (LVLMs) offer a promising alternative through their strong reasoning and generalization capabilities, but fall short on precise, low-level perception tasks such as object detection and accurate bounding box generation. Furthermore, they often require a substantial amount of data for effective fine-tuning on domain-specific tasks. In this paper, we propose a hybrid framework that decouples detection from damage assessment, combining the precision of CV models with the reasoning power of LVLMs. A CV model first detects buildings and generates bounding boxes on the image that are then passed to an LVLM for damage classification and contextual interpretation. We evaluated our framework on two real-world benchmarks: RescueNet and FloodNet. In particular, the best combination under this framework accurately counts intact, partially damaged and completely destroyed buildings, surpassing isolated baselines by up to 2.1 R^2 points, while requiring only limited annotated data for the detection stage. Beyond reporting aggregate gains, we provide a detailed analysis of failure scenarios and edge cases, offering practical insights for practitioners and concrete directions for future work. Our source code and data are publicly available to the research community via the following repository: https://github.com/ungquanghuy-kddi/VLM_GDINO.git
Authors: Mingzhen Xu, Can Xu, Di Wang, Haonan Guo, Bo Du
Abstract: In real-world hyperspectral scenes, pixel representations are often ambiguous due to factors such as spectral similarity, mixed pixels, and local context interference, which may simultaneously encode discriminative evidence and interfering information. Existing methods mainly focus on learning more powerful representations or modeling broader contexts, but rarely investigate whether the learned representations provide reliable evidence or introduce interference into classification decisions. To address this issue, we view hyperspectral image classification from the perspective of pixel-level evidence reliability modeling and propose PNEC-Mamba, a prototype-guided positive-negative evidence calibration framework. The framework progressively establishes semantic references, separates class-related evidence from interference, estimates pixel-level reliability, and performs selective calibration. First, a full-image state-space encoder extracts pixel representations, while dynamic class prototypes provide semantic references that evolve jointly with the feature space. Subsequently, positive and negative evidence is derived from pixel-prototype competition, explicitly separating discriminative cues that support classification from confusing signals associated with competing classes. Based on these evidence relationships, a multi-source uncertainty estimation strategy is introduced to assess pixel-level reliability, enabling stronger evidence calibration for uncertain regions. Finally, a full-resolution consistency refinement step is applied to recover local spatial details and improve boundary coherence in the final predictions. Extensive experiments on three benchmark datasets demonstrate that PNEC-Mamba achieves superior classification performance compared with state-of-the-art methods.
Authors: Wenxuan Ji, Jin Xiao, Xiaoguang Hu, Jiaqi Shi, Zichong Jia, Baochang Zhang
Abstract: Simultaneous localization and mapping (SLAM) based on Neural Radiance Fields (NeRF) enables dense, continuous scene reconstruction. However, existing systems operating with limited online resources struggle to simultaneously construct two types of constraints, namely, compact yet discriminative spatial constraints derived from scene representations and persistent temporal constraints derived from historical observations. To address this challenge, we propose CHOW-SLAM, a dense RGB-D SLAM framework that explicitly constructs these complementary spatial and temporal constraints. Spatially, we propose a compact parametric-hash (P-H) hybrid representation that organizes components based on planes and grids across scales in P and H branches. A unified multi-output decoder further aligns the ray termination distributions induced by TSDF and density, preserving geometry and appearance under a compact parameter budget. Temporally, we propose a complementary overlap-window strategy to prevent optimization from being dominated by short-term overlap or weakly related historical observations. Within a fixed budget, the strategy retains recent frames, selects high-overlap local frames, and introduces temporally distributed historical keyframes. Loss-aware keyframe insertion and bundle adjustment scheduling further adapt optimization to tracking quality. In addition, ORB-based tracking and geometric pose estimation are used for pose initialization, followed by neural rendering optimization to improve tracking stability. Extensive evaluations on multiple datasets demonstrate that CHOW-SLAM outperforms state-of-the-art methods in both scene reconstruction quality and camera tracking accuracy. The source code is available at https://github.com/jinjidexiaohuoban/CHOW-SLAM.
Authors: Wenxiao Fan, Jingling Fu, Fang Li, Luohang Liu, Yu He, Lichen Ma, Zhiyang Yu, Weishan Bi, Junshi Huang, Yan Li, Gu Simiu, Kan Li
Abstract: Vision-language models (VLMs) are expected to revise their reasoning when visual evidence changes. Failures to do so are often attributed to insufficient visual attention or contextual inertia, leaving unclear what models reuse instead of recomputing from the current image. We show that evidence-bearing reasoning in a prior chain of thought (CoT) can form a textual shortcut that competes behaviorally with visual recomputation. Across 16 VLMs, a matched counterfactual analysis identifies evidence-bearing content as the most robust carrier of prior-CoT influence. Removing this evidence-bearing content shifts answer preference more than removing length-matched non-evidence context or the final-answer span, with prior control weakening progressively as more stale evidence is removed. Reordering this evidence also weakens prior control, showing that its organization modulates shortcut strength. Beyond the immediate answer, the shortcut can retain residual influence after answer correction: weakening current-image support shifts preference back toward the prior answer, while repeated prior answers and reused premises arise mainly when the shortcut remains active. To limit this influence, we introduce Fresh-State Attention Firewall (FSAF), a training-free intervention that isolates fresh computation from the prior CoT. Across five VLMs, FSAF raises visual update rate from 35.28% to 53.61% and reduces prior-answer rate from 39.22% to 3.67%. Reliable VLM self-reflection therefore requires more than looking again: fresh visual recomputation must be protected from stale textual reuse.
Authors: Xianjing Han, Yuhan Su, Yang Deng, Dong Ma, Wee Peng Tay, Bin Zhu
Abstract: Text-to-video (T2V) generation models have advanced rapidly, yet their ability to represent diverse cultural contexts remains underexplored. Existing benchmarks mainly focus on perceptual quality, physical plausibility, and text-video alignment, but do not directly assess whether generated videos capture culturally specific objects, actions, rituals, visible text, or audio cues. We introduce CultureVidBench, a comprehensive benchmark for evaluating cultural understanding in T2V generation. CultureVidBench contains 1,000 curated prompts covering 12 countries, 6 continents, 8 cultural regions, and 14 cultural aspects organized into three categories: material culture, social practice & performance, and ritual & ceremony. Designed specifically for video generation, CultureVidBench emphasizes dynamic and multimodal cultural representation, including social interactions, ritual procedure, and culturally appropriate visible text and audio. We evaluate seven representative T2V models through human user studies and MLLM-based automatic assessment across cultural faithfulness, multimodal cultural rendering, semantic adherence, and perceptual quality. Results show that although current models achieve strong semantic adherence and visual quality, they often fail to faithfully capture fine-grained cultural details, particularly for underrepresented regions, rituals, and multimodal cultural cues.
Authors: Liming Tan, Ye Chen, Hao Zhang, Lirong Qian, Feifei Li, Bingbing Ni
Abstract: Controlling human motion and camera movement is essential for faithful human-oriented video generation, yet remains challenging in multi-person scenes with large body motions, occlusions, and dynamic cameras. Existing pipelines typically rely on visual motion sequences, such as skeleton maps, pose maps, or rendered body representations, for motion control, while using camera embeddings for camera control. Such heterogeneous control interfaces force video generation models to reconcile pixel-aligned visual cues with non-visual geometric embeddings, making motion-camera attribution difficult and sensitive to camera estimation errors. We propose \textbf{UniMoCa}, a representation-driven framework that unifies motion and camera controls in visual space. At the core of UniMoCa is \textbf{Motion-Camera Visual Proxy} (\textbf{MCVP}), a mutually-sharable novel representation that converts 3D human motion and camera trajectories extracted from driving videos into an identity-neutral visual proxy. MCVP renders temporally aligned human geometry under the recovered camera trajectory and augments it with explicit camera trajectory markers, replacing heterogeneous visual-parametric controls with distinguishable visual cues. As both control factors are represented in the same visual space, they become mutually compatible rather than heterogeneous, enabling consistent joint reasoning and editing during video generation. We further curate a \textbf{MCVP-Video} dataset covering complex actions, multi-person interactions, and diverse camera trajectories. Experiments based on the Wan2.2 I2V show that UniMoCa achieves substantial gains in human motion control, camera control, temporal consistency, and camera-aware robustness with minimal additional complexity. More details are shown in our Project page: https://tanliming-daniel.github.io/UniMoCa/.
Authors: Cheng-Yao Hong, Ting-Wei Lin, Yun-Chung Lai, Hua-Wei Lee, Hwann-Tzong Chen, Tyng-Luh Liu
Abstract: Long-horizon event-based action understanding remains underexplored because existing datasets largely comprise short, trimmed clips, while collecting native event streams with dense temporal annotations is costly. We introduce Event ActivityNet, a large-scale simulated-event benchmark derived from human-annotated, untrimmed ActivityNet videos. It comprises 3,263 videos, 200 action classes, and 106.94 hours, with matched 5-bin and 9-bin event-voxel representations, temporal action annotations, and timestamped captions. The benchmark supports annotated-segment action recognition, auxiliary event-language alignment, and causal online temporal action localization. We generate event voxels directly from non-interpolated source videos in decoded frame order, retain per-video rational nominal or average frame-rate metadata for approximate time mapping, and use action-center reconstruction LPIPS as a soft diagnostic of retained reconstructable content. We establish baselines for adaptive event framing, prompt-caption alignment, and event-only, RGB-only, and RGB-event localization. Under a progressive nested-scale training protocol, recognition Top-1 accuracy increases from 52.25 to 66.42, while online temporal localization average mAP improves from 21.7 to 29.0. Moreover, staged Event ActivityNet pretraining followed by native-event fine-tuning consistently outperforms target-only and joint-from-scratch training across multiple supervision budgets. Event ActivityNet provides a scalable benchmark for long-horizon event modeling, although native-camera evaluation remains essential for deployment-oriented conclusions.
Authors: Lingwei Dang, Shishuo Shang, Pan Liu, Jiajia Cheng, Ziyan Qiu, Zhenhao Zhang, Yufei Zhu, Shenghui Huang, Qingxin Xiao, Yun Hao, Juntong Li, Qingyao Wu
Abstract: Fixed-layout indoor furniture styling requires selecting assets that form a coherent room without changing the prescribed furniture categories, positions, orientations, or scales. Existing approaches typically retrieve each asset independently or rely on static local relations, making them prone to shape, material, and color conflicts after scene composition. We introduce StyleForge, a scene-level structured selection framework built on a dynamic hypergraph style field. A frozen multimodal large language model extracts structured style priors from an open-ended style request and the fixed layout, while StyleForge maintains a learnable candidate distribution for each furniture slot. Conditioned on the target style, the dynamic hypergraph style field adaptively activates and weights layout-induced hyperedges to capture higher-order dependencies among furniture. Counterfactual style preference learning then treats each candidate as a local substitution in the current style field and evaluates its contextual compatibility using Mahalanobis energies. Training alternates between optimizing the style field and the candidate logits. At inference, the model remains frozen and test-time training updates only room-specific candidate logits, progressively correcting cross-slot style conflicts as the global scene context evolves. Experiments on 3D-FRONT demonstrate state-of-the-art furniture retrieval and scene-level style coherence, producing more coherent fixed-layout furniture arrangements than object- and scene-level retrieval baselines.
Authors: Zhengyang Zhang, Ziyu Lu, PengCheng Li, Hongbo Duan, Yi Liu, Pengting Luo, Peiyu Zhuang, Xinghui Li, Shaohua Ma
Abstract: 4D Gaussian Splatting (4DGS) excels in dynamic 3D reconstruction and real-time novel view synthesis via efficient 4D Gaussian representations and parallelizable rendering. However, existing 4DGS approaches rely on a single polynomial to model motion, which limits performance in complex dynamic scenes where high-frequency motion components are prevalent, and fails to ensure long-term stability due to cumulative trajectory drift. To address these issues, we propose a Fourier Motion Modeling module: this paradigm decomposes motion into frequency-based sinusoidal components, capturing both low-frequency global trajectories and high-frequency local details to model complex motion patterns accurately. It retains the real-time rendering capability of 4DGS while improving complex motion fitting and long-term coherence. Additionally, we integrate a motion-aware regularization strategy into the loss function: it uses frequency-dependent weights to suppress high-frequency jitter while preserving low-frequency motion coherence. Extensive experiments on N3V and Google Immersive datasets from multiple scenarios demonstrate the effectiveness of our method.
Authors: Horst Hammer, Sylvia Hochstuhl, Antje Thiele, Tobias Brosch, Padraig Davidson, Tim Remiger, Michael Teutsch
Abstract: Ship detection in Synthetic Aperture Radar (SAR) images plays an important role for maritime situational awareness, especially with respect to different illegal activities at sea such as illegal fishing, smuggling or border violations. Modern ship detection methods using neural networks usually require large training datasets, which are considerably scarcer in the SAR domain than in the electro-optical domain. While several free datasets exist for this task, their availability and usability vary. In this paper, OpenSARShip-Ship Detection Dataset (OSSDD), a new dataset based on the well-known OpenSARShip 1.0 dataset is proposed for training neural networks for SAR ship detection. OSSDD is freely available and contains 15,197 Sentinel-1 amplitude patches in VV and VH polarization, binary ship masks, axis-aligned bounding box and rotated bounding box annotations for a total of 55,759 ships. The construction of the dataset, the contents and structure of the downloadable data and experiments with three common detector models (Faster R-CNN, FCOS, DETR) are shown and discussed. The results serve as benchmarks for future experiments. The dataset is available on Hugging Face at https://huggingface.co/datasets/sylviaHoch/OpenSARShip-Ship-Detection-Dataset.
URLs: https://huggingface.co/datasets/sylviaHoch/OpenSARShip-Ship-Detection-Dataset.
Authors: Ziyu Ma, Hailang Huang, Shun Zou, Yong Wang, Shidong Yang, Yiming Hu, Fei Wei, XiangXiang Chu
Abstract: Large language model (LLM) agents increasingly undertake long-horizon tasks that require sustained reasoning, tool use, and revision across many interdependent steps. However, existing agent harnesses maintain task execution, task state, and completion assessment within a growing context, making the state difficult to track and allowing incorrect self-assessments to propagate into later decisions. We reformulate long-horizon execution as a task-state management problem and propose LongHorizon-Harness, which maintains the task state explicitly outside execution and updates it only with facts independently verified from the environment. Its Manage-Execute-Audit(MEA) loop uses a manager to maintain the task state and determine the next subtask, a fresh-context executor to perform it, and a read-only auditor to verify the resulting environment state before the next round. A lightweight AgentAdapter supports interchangeable model and harness backends without modifying their native agent loops. LongHorizon-Harness improves Qwen~3.7-Plus from 51.8% to 80.7% on WeaveBench, from 69.7% to 77.2% on Terminal-Bench~2.1, and from 2.8% to 8.3% on OSWorld~2.0. It also raises Claude Opus~4.7 from 20.0% to 34.3% on an OSWorld2.0 subset, demonstrating consistent gains across models, harnesses, and interaction domains.
Authors: Yiming Wang, Ye Chen, Hanqi Chen, Bingbing Ni
Abstract: Multimodal large models are increasingly used to generate scalable vector graphics (SVG), but reliable evaluation remains underexplored. Existing protocols are often code-centric or borrow raster-image metrics after rendering SVGs, which fail to reflect human perception and overlook SVG-specific qualities such as geometry and spatial composition. We introduce SVGEval, a vision-grounded multimodal benchmark for human-aligned SVG quality assessment. SVGEval explicitly incorporates visual renderings to evaluate whether models can judge the rendered outcome rather than only inspect SVG code, and provides high-quality annotations obtained via multi-round human labeling with expert refinement. Systematic evaluations across representative multimodal models reveal a clear gap: models perform relatively well on semantic alignment and aesthetics, yet struggle on geometry- and layout-related judgments. Building on SVGEval, we train an explainable SVG quality scorer that outputs multi-aspect scores with textual rationales. Ablations show that explicit visual grounding and reasoning supervision are crucial, especially for spatial and geometric assessment. SVGEval offers a reliable testbed and practical scorer for evaluating and improving SVG generation in the era of multimodal models.
Authors: Haijie Yang, Jindi Bao, Yixuan Dong, Hongliang Zhang, Jian Bi, Hao Tang, Zhenyu Zhang, Jianjun Qian, Jian Yang
Abstract: Audio-driven portrait animation has advanced rapidly with diffusion-based generative models, yet real-time one-shot generation with expressive emotion control remains challenging. Existing methods often suffer from insufficient emotion-aware motion priors and expensive appearance computation during multi-step denoising. To address these issues, we propose Proxy Avatar Meets Low-Rank Caching, a cascaded framework for real-time one-shot emotion-controllable portrait animation. Instead of directly generating the target portrait from audio, our method uses a Gaussian-based emotion proxy avatar as a reusable motion generator, which is trained once on a single identity to produce expressive driving videos from audio and emotion labels. Since the proxy avatar only provides motion rather than target appearance or geometry, a large-scale one-shot retargeting model further extracts identity-independent motion from the proxy performance and adapts it to arbitrary target portraits. To improve inference efficiency, we introduce zero-shot appearance reuse with low-rank caching, which caches reference appearance features at the initial denoising step and models subsequent feature variations using lightweight low-rank adapters. Extensive experiments demonstrate that our method achieves stronger emotional expressiveness, better identity-preserving animation, and substantially reduced inference cost, enabling real-time one-shot portrait animation.
Authors: Zizhong Ding, Junxian Li, Kai Liu, Shaoqiu Zhang, Xiao Xiao, Linghe Kong, Yulun Zhang
Abstract: Visual token pruning reduces the inference cost of multimodal large language models, but a fixed token ratio is poorly matched to text-rich inputs. In OCR-centric tasks, decisive evidence can be a small number, label, or field whose relevance is specified by the question; indiscriminate pruning can erase that evidence while retaining visually salient but irrelevant regions. We present ET-Prune, a training-free framework that casts pruning as evidence allocation. It derives question-conditioned evidence from a decoder-side partial query-key block, safeguards text-like spatial regions, and converts evidence uncertainty and density into a sample-specific token floor. Three progressive middle-layer events then move the sequence toward this budget, retaining more tokens for diffuse or text-dense evidence and pruning concentrated evidence more aggressively. At the observed point estimates from one deterministic pass per configuration, ET-Prune leads or ties among pruned methods in all six backbone-benchmark comparisons at roughly half tokens. On OCRBench-v2, it leads the strongest pruned baselines by 1.80 and 0.68 percentage points on Qwen3-VL-8B and InternVL3.5-8B, respectively, while retaining about half of the visual tokens; on MMBench v1.1, it reaches 0.8467 circular exact-matching accuracy versus 0.8437 for Vanilla at 54.45% average visual-token retention. These results show a favorable observed quality-cost trade-off for evidence-aware dynamic budgeting in text-rich multimodal inference.
Authors: Jingqi Tian, Haoji Zhang, Lin Chen, Hongbo Jin, Haonan Xu, Tianrui Zhu, Xingming Shui, Shilin Ma, Wenjing Yang, Yansong Tang
Abstract: Chain-of-thought (CoT) reasoning can improve performance on difficult video questions but often wastes decoding tokens on simple ones. We study whether a video multimodal large language model can adapt its reasoning effort to each question. We propose AdaThinkV, an adaptive framework for video reasoning that learns whether to reason explicitly without offline difficulty labels, manually tuned confidence thresholds, or an external router. During reinforcement learning, AdaThinkV samples matched rollouts in explicit reasoning and direct answering modes for each prompt. ThinkGain estimates the prompt-level utility of explicit reasoning by balancing its accuracy gain against additional response length, providing supervision for both conditional response generation and autonomous mode selection. For difficult prompts, limited rollout exploration can yield groups in which every response is unsuccessful and accuracy rewards show little variation, providing insufficient signal for learning. We therefore introduce Variance Recovery Policy Optimization (VRPO), which retains and progressively expands these groups to recover informative signals from prompts that are difficult yet solvable. At inference, AdaThinkV selects a response mode and generates the response in a single autoregressive sequence. Across a unified suite of video reasoning evaluations, AdaThinkV achieves a mean accuracy of 40.79 with an average of 257.20 output tokens, outperforming the strongest evaluated adaptive baseline by 2.98 points while using 22.7% fewer tokens. Project page: https://trilarflagz.github.io/AdaThinkV/
Authors: Landi He, Mingde Yao, Shawn Young, Lijian Xu
Abstract: Visual token pruning reduces the computational cost of Vision-Language Models (VLMs) by removing redundant visual tokens. The key is to learn a score that measures whether a token is useful. Existing methods typically rely on Gumbel-Softmax to approximate discrete selection during training. Such selectors make the score depend on the behavior of a relaxed pruning operator, not directly on the consequence of information loss. In this paper, we propose DiffPrune, which gives token scores a direct meaning. During training, DiffPrune keeps all tokens and weakens each token's information according to its score. If weakening a token hurts the task, the scorer is pushed to protect it; if not, the token can receive a lower score. Because the loss is differentiated through this actual information-throttling path, the scorer avoids the unstable surrogate path of relaxed token selection. DiffPrune implements this idea with an Information Throttler, which injects variance-preserving noise into visual tokens, where high-score tokens remain close to their original representations, while low-score tokens carry less original information. At inference, the throttler is removed, and hard top-K pruning is applied using the learned scores. Across ten VLM benchmarks, DiffPrune retains 96.5% of full-model accuracy while accelerating LLM prefill by 2.85x, with only 0.69 ms inference overhead. Code will be publicly available.
Authors: Kun Guo, Yuzhou Yang, Haoyue Wang, Qichao Ying, Sheng Li, Zhenxing Qian
Abstract: Rapid advances in image generation models call for interpretable AI-generated image detection methods that not only determine authenticity but also provide supporting visual evidence. Existing approaches may produce inconsistencies between generated explanations and localized evidence regions, undermining the reliability of explanations for authenticity decisions. Meanwhile, existing benchmarks provide limited coverage of the diverse human-centric scenes prevalent in generated imagery. To address these limitations, we investigate authenticity detection with grounded and explainable visual evidence in human-centric scenes. We present HAVE (Human-centric AI-generated Visual Evidence), a diverse human-centric dataset comprising 40K real and 39K AI-generated images from 10 recent generators, with 106K localized evidence instances across 8 evidence categories, each annotated with a bounding box and a region-aligned explanation. We further propose PAVE, a Perception-Aware Visual Evidence framework that jointly performs authenticity prediction, visual evidence grounding, and region-aligned explanation generation. PAVE employs a judge-guided alignment reward to assess region--explanation consistency and evidence validity, together with perception-aware regularization that contrasts token-level predictions between original and randomly masked images to promote reliance on visual input. Experiments on HAVE and external datasets demonstrate strong performance in authenticity detection, visual evidence grounding, and explanation quality. Code and data will be released upon publication.
Authors: Zong-Wei Hong, Jinglun Li, Shen Zhang, Yuhan Liu, Linze Li, Yao Tang
Abstract: Denoising diffusion transformers achieve strong generation quality but converge slowly during training. Regularizing their internal representations has emerged as an effective accelerator, yet existing methods split into two families with complementary costs. Target-based methods strengthen representations by aligning them to external features, which requires an external encoder and a learnable projection head to bridge feature spaces. Target-free methods hold no reference at all, and can only repel the model's own features across samples or layers, discarding whatever structure the data contains. Prior work suggests that spatial structure, rather than global semantics, drives the gains of alignment. We therefore ask whether such structure can serve as a target directly, and whether it exists not only within an image but across images. Our key insight is that the clean data latent already carries this structure in the relations among its tokens, where a relation is the similarity between two tokens, a single scalar comparable across feature spaces without a projection head. We propose Structural Parameter-free Affinity Regularization (SPARE), a regularizer that matches the pairwise affinities of intermediate tokens to those of the clean latents. To exploit this structure fully, SPARE extends the matching to token pairs across images, precisely the pairs that prior target-free methods repel by default, and calibrates both relation types with a single learning objective. On ImageNet $256 \times 256$ with SiT backbones under matched 400K-iteration budgets, SPARE adds no encoder, head, or parameters and only 0.08 GB of training memory, yet attains the lowest FID among parameter-free regularizers in every tested setting, recovers 37 to 54\% of REPA's FID reduction, and improves over REPA when combined with it, reaching FID 1.90 under classifier-free guidance at 1M iterations.
Authors: Junyu Zhu, Hao Zhu, Xinzhuo Zhang, Hongdong Li, Zhan Ma, Xun Cao
Abstract: Dynamic 3D scene reconstruction has achieved remarkable success under the assumption of strictly synchronized multi-camera inputs. However, in real-world scenarios, temporal asynchrony among capturing devices remains a critical challenge, leading to severe motion blur and geometric artifacts. Existing asynchronous reconstruction methods typically estimate temporal offsets through photometric supervision, but appearance matching provides weak temporal cues under large offsets and complex motions. We attribute this limitation to two major bottlenecks: texture-induced collapse, where low-texture regions provide nearly vanishing alignment signals, and deformation-induced coupling, where temporal errors are absorbed into distorted geometry or motion rather than being explicitly corrected. To address these issues, we propose ASTRA (Asynchronous Spatio-Temporal Reconstruction via Trajectory Alignment), a framework that introduces 2D motion trajectories as explicit, texture-agnostic supervision for asynchronous dynamic reconstruction. Instead of synchronizing cameras solely through rendered color residuals, ASTRA jointly optimizes temporal offsets and dynamic 3D representations by aligning the projected motion of reconstructed 3D points with observed 2D trajectories, while using dynamic and certainty masking to suppress unreliable trajectory constraints. Extensive experiments on different dynamic Gaussian Splatting backbones show that ASTRA preserves high-frequency spatial details and sustains strong robustness even under severe asynchrony with up to 25-frame offsets, achieving approximately 1.4 dB PSNR improvement, reducing temporal-offset MAE by 54.0\%, and nearly quadrupling the synchronization success rate.
Authors: Kaiyi Zhang, Zhihao Liang, Haolin Liu, Qingxiang Lin, Zeqiang Lai, Yunfei Zhao, Bowen Zhang, Xianghui Yang, Zibo Zhao, Chunchao Guo, Long Quan
Abstract: Sparse voxel grids preserve the spatial structure needed for detailed 3D reconstruction, but their memory still grows rapidly with resolution as active surface cells increase. We introduce ChunkVAE, a sparse grid variational autoencoder organized around local chunks rather than a global latent volume. Local learned operators permit independently chosen encoder and decoder partitions and allow inference chunk sizes to differ from training. Two complementary data operators make this flexibility practical: Balanced Binary Object Partitioning distributes active cells while limiting replicated overlap, while S-Curve weighted stitching attenuates unreliable boundary features when assembling a global latent or reconstruction. Across three object benchmarks, ChunkVAE is competitive with or better than strong baselines from $512^3$ to $1536^3$; smaller chunks lower peak allocated memory and shorten per-chunk compute, enabling faster parallel inference. Stable stitched latents and improved image to 3D metrics indicate that local compression can scale geometry while retaining the global interface required downstream.
Authors: Jia-Chen Zhang, Ze-Yu Zhang, Kai-Wei Zhang
Abstract: Computer-use agents (CUAs), which empower large language models to autonomously operate operating systems and the web, are increasingly vulnerable to indirect prompt injection attacks. A widely adopted defense is the human-in-the-loop paradigm, in which the agent pauses for explicit user confirmation before executing sensitive operations. While effective against conspicuously high-harm attacks, this defense offers little protection against what we term Invisible Ink Threats: low-harm injected goals, such as starring a repository or installing a package, that are behaviorally indistinguishable from legitimate task execution and thus evade both model safety mechanisms and human oversight. To systematically investigate this blind spot, we present II-Bench, a collection of seemingly harmless adversarial tasks. II-Bench comprises 444 examples targeting confidentiality and integrity attacks across three platforms, spanning three attack categories: page navigation and interaction, sensitive information exfiltration, and code download and execution. Each category is instantiated in both natural language and code forms under two levels of instruction specificity. Furthermore, we construct HITLCUA, a comprehensive adversarial testing framework that integrates a real virtual machine operating system environment with isolated Docker-based web platforms, and simulates human participation by allowing CUAs to consult an API-simulated user before proceeding with suspicious operations. Extensive evaluations of leading CUAs reveal that low-harm injections frequently bypass both agent defenses and simulated user review, exposing severe and previously underexplored security risks in current CUAs.
Authors: Hongjie Zhou, Shiqin Wang, Haoyang Chen, Haonan Guo, Di Wang, Juhua Liu, Fu Lin, Yong Luo
Abstract: Remote-sensing videos enable real-time observation of changes in target attributes, short-term activities, and scene evolution. They record motion, actions, interactions, and scene changes that cannot be captured by isolated images. Existing models primarily target single images or discrete temporal observations spanning a long time range. However, a unified evaluation setting for assessing vision-language models on continuous remote-sensing video understanding remains lacking. We introduce RSVideo-10K, a remote-sensing video dataset comprising 10,773 instances, 1.47 million frames, and 17.02 hours of footage, containing both unmanned aerial vehicles and satellite platforms. Its fixed evaluation benchmark, RSVideo-Bench, contains 2,731 test instances and evaluates two complementary aspects of remote-sensing video understanding: L1 Perception and L2 Reasoning, spanning seven capability groups and 17 tasks. Evaluations show that current vision-language models still struggle to recover small local evidence, track short-lived states, and use scene-constrained spatial relations. Based on this analysis, we further propose RSVideo, a reinforcement learning framework for small-target spatiotemporal focusing that selects question-relevant regions across frames and suppresses redundant background tokens. RSVideo achieves a maximum absolute improvement of 9.01% with InternVL3.5-14B and attains the highest accuracy of 40.63% with Qwen3.6-27B across 26 open-source vision-language backbones. Codes will be available at https://github.com/HongjieZhou0329/RSVideo.
Authors: Haofan Cao, Zhichao You, Yunkai Yang, Liang Guo, Jie Wang, Chongshou Li
Abstract: Tracking links observations of the same object through visual change, yet cannot by itself determine when the object is empty or filled, intact or cut. We formulate identity-conditioned state-moment retrieval: given a tracked-object history and alternative state descriptions, localize an interval in which each described state holds. Absolute image-text similarity scores descriptions independently; because every visible frame depicts the same target, shared object compatibility can obscure the state evidence needed to identify the target interval. The alternatives provide the missing reference: evidence for one state should be measured against the others. We introduce D\'ej\`a Cue, a training-free framework that turns these alternatives into a vocabulary-relative coordinate system. It subtracts their state-balanced centroid from each description, calibrates frame scores, and scans multiple durations within contiguous visible runs using a frozen encoder. On 78 VOST histories, holding the temporal scan fixed and changing only the query reference nearly doubles R@1 at tIoU 0.5 from 10.3\% to 20.5\% and raises Top-1 tIoU from 16.0\% to 21.5\%. Candidate-rank analyses show that vocabulary-relative queries rank useful intervals higher within the same candidate set. Related state descriptions can therefore serve as an object-specific, query-time coordinate system for reading frozen visual representations.
Authors: Ji Huang, Yongsheng Dai, Tianyu Ren, Barry Devereux, Hui Wang
Abstract: Recent advances in proposal-free Video Moment Retrieval (VMR) have highlighted the effectiveness of Static Scene Graphs (SSGs). By modeling objects and their relations at the frame level, SSGs enrich retrieval-oriented video representations. However, integrating SSGs into VMR remains constrained by two inherent limitations: (1) Lack of Temporal Dynamics. SSGs fail to model how objects and their relationships evolve over time, leading to the loss of essential temporal dependencies in video representation; and (2) Lack of Explicit Temporal Span Encoding. SSGs do not explicitly encode the duration of relationships, making precise localization challenging. To address these limitations, we propose Temporal Bipartite Scene Graph Network (TBSG-Net)---to the best of our knowledge, the first Dynamic Scene Graph (DSG) based proposal-free VMR model. Specifically, TBSG-Net leverages DSGs to extract event-centric graph representations of the input video, enabling the modeling of object interactions over time and thus addressing limitation (1). These DSGs are then processed by a novel Dynamic Scene Graph Embedding (DSG-E) module to capture both Temporal Span and spatio-temporal information. First, DSG-E utilizes a TBSG Constructor to transform DSGs into TBSGs, explicitly encoding objects, relationships, and time spans to tackle limitation (2). Second, the resultant TBSGs are passed into a hybrid TBSG Encoder that integrates a Transformer variant for global event modeling and a Graph Convolutional Network for detailed relational reasoning, ultimately producing a more comprehensive spatio-temporal representation. Our experiments demonstrate substantial improvements of TBSG-Net over all baselines.
Authors: Zitong Xu, Huiyu Duan, Xinyun Zhang, Weifei Xiong, Tianyi Zheng, Xiongkuo Min, Qiang Hu, Zhengxue Cheng, Bo Li, Guangtao Zhai
Abstract: Recent advances in unified multimodal models have significantly improved text-guided image editing abilities. In particular, models such as Nano-Banana-Pro and GPT-Image-2 demonstrate emerging capabilities in multi-source image editing (MIE), including tasks such as object synthesis, person-background composition, and cross-image style fusion. However, existing benchmarks and image editing assessment (IEQA) methods remain primarily focused on single-image editing tasks and largely overlook the more challenging setting of MIE. This highlights the urgent need for a comprehensive and human-aligned benchmark for MIE. To this end, we introduce MIE-Bench, the first large-scale multiple image editing benchmark with fine-grained human preference annotations. Specifically, MIE-Bench includes 3,000 editing instances across 16 tasks, each involving more than two source images and an editing prompt, together with 36K edited images produced by 12 state-of-the-art editing models and over 108K mean opinion scores (MOSs) covering visual quality, instruction following, and attribute preservation. Based on MIE-Bench, we propose MIEScore, a multimodal large language model (MLLM)-based evaluation model enhanced with skill optimization and multi-dimensional supervised fine-tuning, to provide human-aligned feedback for MIE. Extensive experiments show that MIEScore achieves state-of-the-art performance in aligning with human preferences and generalizes well across other IEQA datasets. Both the dataset and the model are available at https://github.com/IntMeGroup/MIEScore.
Authors: Xianghui Fan, Zhaoyu Chen, Bingqian Wu, Dayu Li, Xin Zeng, Huanran Cui, Guangzhen Xu, Xiangru Huang, Hang Yang
Abstract: Monocular depth foundation models, benefiting from large-scale synthetic training data, have demonstrated strong generalization. However, they often hallucinate depth on non-Lambertian surfaces, estimating reflected content in mirrors or transmitted content behind glass rather than the physical surface itself. Adapting these models with real-world data is challenging because conventional depth sensors are also unreliable in such regions. We observe that while the appearance of a non-Lambertian surface varies with its reflected or transmitted environment, its underlying geometry remains unchanged. Based on this observation, we propose GIFT (Geometry-Invariant Fine-Tuning), a parameter-efficient post-training framework that requires no measured depth labels. We collect groups of RGB images under controlled appearance changes while keeping the camera and target geometry fixed. GIFT exploits geometric invariance across these observations to suppress non-Lambertian depth hallucinations while retaining general depth estimation capability. We further construct a controlled benchmark that evaluates non-Lambertian depth recovery, robustness to appearance changes, and performance retention in other regions. Experiments on our benchmark and an independent real-world dataset demonstrate that GIFT improves depth prediction for mirrors and transparent objects while largely preserving the base model's performance, providing a practical and low-cost approach for adapting monocular depth foundation models to non-Lambertian scenes.
Authors: Zhu Chen, Dingkun Liu, Yuheng Chen, Dongrui Wu
Abstract: Brain-computer interfaces (BCIs) have been widely used in motor rehabilitation, disease diagnosis, and other neural engineering scenarios. However, conventional neural signal decoding algorithms often suffer from limited generalizability and high adaptation costs, motivating recent interest in BCI foundation models. Existing approaches still struggle to jointly achieve general transferability, accurate decoding, and efficient downstream adaptation. We present STEAM, a hierarchical transfer framework that reconciles general-purpose representation learning with paradigm-specific specialization in EEG foundation models. The framework is instantiated as a dual-branch spatio-temporal encoder in which a shared soft mixture-of-experts (SSMoE) module aligns the spatial and temporal branches, allowing complementary representations to exchange information through a compact set of soft slots. Across seven downstream datasets and fourteen evaluation settings, STEAM attains the best average rank among the compared methods at a competitive inference cost measured in FLOPs. Building upon the Stage-I general initialization, the hierarchical pre-training strategy further specializes the model to a target paradigm without retraining from scratch, yielding consistent gains in paradigm-specific decoding accuracy.
Authors: Guandi Wang, Ming Li, Yunsen Xing, Junle Liu
Abstract: Deep multimodal fusion for object detection has demonstrated good performance through mining modal characteristics. However, existing feature-level fusion methods mainly weigh between two modalities and unify them in a unified representation space. This can lead to overfitting or over-specialization of the statistical properties of a single modality within a dual-backbone architecture. This paper proposes an Attention-Driven Complementarity Resampling framework for robust improvement of cross-modality object detection. Based on a shared channel spatial attention mechanism, we first introduce the semantic mask exchange to actively mix the boundaries of the modalities during the training phase, forcing the backbone network to learn generalized features without relying on fixed modal labels. Then we propose a learnable channel competition to sample and aggregate features in a channel-wise and learnable way. Our experiments on multiple datasets demonstrate that the proposed method is effective and yields competitive results among existing state-of-the-art approaches. The source code is provided in the supplementary material.
Authors: Tianyu Liang, Xiangxi Zheng, Yilin Wang, Dongxing Mao
Abstract: Vision-Text Compression (VTC) renders long texts into images and encodes them through the vision encoder (ViT), compressing thousands of text tokens into far fewer visual tokens. However, since the ViT is pretrained predominantly on natural images, it captures visual attributes (glyphs, font sizes, layout) rather than linguistic semantics, causing rendered-image representations to diverge from native-text representations. We term this cross-path inconsistency and show, via rendering perturbation experiments, that it is a critical yet overlooked bottleneck of VTC. We propose SPIRAL (Self-improving Path Integration and Realignment), a self-supervised alignment framework that closes this gap using only the model's own text-path behavior as supervision, requiring no external teachers or additional annotations. SPIRAL operates at two complementary granularities: token-level on-policy distillation (OPD) for local faithfulness, and sequence-level preference optimization (DPO) for global coherence. On VTCBench, SPIRAL improves the overall score of Qwen3-VL-8B from 35.10 to 54.02, approaching the native text-input performance (55.60) and outperforming models up to 30x larger. The two granularities exhibit complementary strengths: OPD excels at retrieval and is sample-efficient, while DPO is stronger on reasoning and memory and scales better with data. SPIRAL's benefits also generalize to out-of-domain benchmarks, confirming that effective VTC hinges on aligning rendered-image representations back to native-text semantics.
Authors: Jin Cui, Chuanchang Su, Jiayi Lu, Xinyue Long, Boran Zhao, Pengju Ren
Abstract: Vision-language models (VLMs) remain unreliable when predictions require fine-grained visual evidence. We identify a previously overlooked cause: spectral response rigidity. Despite substantial frequency variation across images and tasks, pretrained vision encoders exhibit persistent, encoder-specific layerwise spectral profiles that change only marginally under downstream fine-tuning. Since pretrained vision encoders only receive images, they cannot adapt spectral extraction to the evidence required by the current query. We therefore propose HAFI-VLM, which introduces a task-conditioned frequency pathway while preserving the pretrained semantic representation. Hierarchical Adaptive Frequency Injection (HAFI) retrieves complementary low-, mid-, and high-frequency evidence at multiple encoder depths using text-modulated, spatially aligned cross-attention. A Visual Enrichment Layer Adapter further recalibrates shallow LLM attention to effectively utilize the enriched visual tokens. Experiments on LLaVA-1.5 and Qwen2.5-VL demonstrate consistent improvements in general VQA, text-rich understanding, and hallucination robustness, outperforming representation-level enhancement methods and most resolution- or cropping-based approaches without additional high-resolution encoding. Mechanistic analyses show that HAFI restores task-dependent spectral allocation while retaining semantic attention, establishing frequency enrichment as a distinct and effective route for improving VLM perception.
Authors: Hangrui Zhang, Feifei Shao, Yawei Luo, Ping Liu, Jiaxiang Liu, Zuoqi Tang, Zhao Wang, Hongwei Wang, Jun Xiao
Abstract: In-context learning (ICL) has attracted increasing attention for enabling models to perform new tasks using only a few ``input--output'' prompt examples. However, existing approaches suffer from \textbf{shallow task adaptation}, where prompts are primarily used as contextual cues to implicitly infer task intent through semantic representations, while the underlying computational process remains unchanged. This limitation restricts task-specific adaptation and compromises inference interpretability. We argue that prompts should not only condition feature representations but also dynamically regulate the model's computation pathways. To this end, we propose \textbf{PromptPath}, an adaptive ICL framework that enables computation-level adaptation through prompt-conditioned dynamic pathways. Specifically, PromptPath introduces a prompt-driven routing mechanism to selectively activate and compose lightweight low-rank experts, forming task-specific computational pathways tailored to different prompts. By integrating prompt information directly into the inference process, PromptPath dynamically reconfigures model computation to enhance task specialization and interpretability. Extensive experiments on 3D point cloud and 2D visual recognition benchmarks demonstrate that PromptPath consistently outperforms state-of-the-art ICL baselines while exhibiting strong cross-domain and cross-task generalization.
Authors: Long Qian, Jiaqi Wei, Bingke Zhu, Yingying Chen, Jinqiao Wang
Abstract: Modern vision language models (VLMs) turn high-resolution images into long sequences of visual tokens. Every token traverses the language decoder and persists in its prompt KV cache, inflating inference cost and motivating aggressive visual compression. Existing score-based methods assign each token an independent importance score and retain the Top-K. However, text queries consume collective, signed attention messages from the visual population, not isolated patches. Consequently, equally sized Top-K sets can repeatedly cover one salient region, omit sparse but complementary evidence and discard information carried by the removed population. We therefore formulate faithful visual compression as constructing a compact coreset for decoder messages, and introduce our training-free Grounded Message Coreset Pruning (GMC) which jointly allocates support across query-grounded, appearance, and coordinate-aware evidence, then transports discarded states into selected representatives at their original multimodal positions before physical compaction and native attention resume. This decomposes faithful compression into two coupled components, including selecting carriers that cover the required message modes and realizing the signed population message on those carriers. We further derive bounds connecting their errors to signed-message distortion, visual innovation, and candidate-margin stability. Experiments across multiple VLM families and diverse benchmarks demonstrate strong performance, with GMC-H2 retaining 97.78% Full-relative mean capability on Qwen2.5-VL-7B using 80.2% fewer visual tokens, while GMC-L16 reaches 100.36%. Controlled interventions verify that collective support and population realization jointly drive these gains.
Authors: Xuanhui Lin, Junhao Dong, Mingrong Gong, Yucheng Chen, Xinghua Qu, Yew-Soon Ong
Abstract: Vision-language models (VLMs) exhibit strong generalization across multimodal tasks but remain vulnerable to adversarial perturbations. Existing attacks typically follow single-trajectory gradient optimization or task-specific objectives, limiting search-space exploration and cross-task transferability. We propose an evolutionary-computation-guided cross-modal attack framework for unified VLMs. The framework adaptively searches both textual and visual spaces. On the textual side, it evolves hard negative semantic embeddings around the source-category representation to provide diverse cross-modal repulsion. On the visual side, it maintains a population of object-region perturbations and combines momentum-based gradient updates with evolutionary selection, mutation, and crossover to more reliably explore multiple feasible trajectories. Jointly optimizing semantic negative guidance and localized perturbations generates adversarial examples that consistently shift source-object semantics toward target categories across vision-language tasks. Theoretical analyses show that the co-evolutionary search preserves perturbation feasibility, prevents degradation of the best observed fitness, and increases the probability of reaching high-margin adversarial regions compared with single-trajectory optimization. Experiments on Florence-2, OFA, and UnifiedIO-2 demonstrate strong overall attack performance across image captioning, object detection, region categorization, and object localization. Ablation studies further verify the complementary effectiveness of text-side semantic evolution and image-side perturbation evolution, as well as the framework's efficiency and cross-task transferability.
Authors: Shivani Mall, Swarnim Jain, Joao F. Henriques
Abstract: Much of the recent progress in image and video recognition has come at the cost of memory: larger models, increased resolution, and longer temporal contexts. An inevitable component is the quadratic (or larger) growth of memory and compute based on image resolution, which is a property of the grid sampling used in convolutional networks and vision transformers. In this work we study residual networks whose convolutional blocks have logarithmic-square growth instead, enabling them to process very high-resolution video quickly. The key insight is to use a residual architecture's residual stream as a high-resolution buffer, to which convolutional blocks only read and write via log-polar image warp operations. Layers adaptively focus on different parts of each frame, with very high resolution only near the focus point. A complete high-resolution representation is built up in the residual stream, analogous to eye saccades creating a complete picture in biological vision, and a theoretical construction is presented that eliminates the quadratic dependency of the residual stream resolution. Experiments demonstrate that our proposed HiResNets learn to foveate around scenes similarly to human vision, and have superior performance in difficult egocentric video recognition tasks, especially egocentric video with small objects and fine-grained recognition.
Authors: Yanqing Song, Jifei Miao, Chaoqian Li, Rui Mei, Kit Ian Kou, Liqiao Yang
Abstract: Division-of-focal-plane (DoFP) color polarization cameras enable snapshot acquisition of color polarization mosaic images, but the inherently sparse sampling pattern makes color polarization demosaicking severely ill-posed. Existing methods often fail to jointly exploit the correlations among polarization channels and the physical constraints inherent in polarization imaging, resulting in noticeable demosaicking artifacts. To address this issue, a quaternion-tensor-based color polarization demosaicking (CPDM) method incorporating Stokes-domain total variation (TV) regularization is proposed. Correlation analysis shows that the correlations among polarization channels are stronger than those among color channels. Accordingly, the color polarization images acquired at $0^\circ$, $45^\circ$, $90^\circ$, and $135^\circ$ are encoded into the four components of a third-order quaternion tensor, with the color channels organized along its third mode. A low-rank prior is then imposed on the quaternion tensor to exploit the global structural redundancy in the color polarization data. Moreover, spatial gradients are mapped to the Stokes domain through an orthogonal transformation to separate intensity, polarization and residual variations, with adaptive quaternion weights enabling component-specific regularization and preserving the energy consistency of the reconstructed Stokes vectors. An efficient optimization algorithm is derived for the resulting model. Extensive experiments demonstrate the superior demosaicking performance of the proposed method.
Authors: Haixu Song, Xiaoke Yang, Shengjun Zhang, Jiwen Lu, Yueqi Duan
Abstract: In this paper, we propose UniqueSplat, a view-conditioned feed-forward 3D Gaussian Splatting model to reconstruct customized 3D radiance fields for each view query. Existing feed-forward methods such as pixelSplat and MVSplat aim to generate fixed Gaussians across all views of each scene by minimizing the error between rendered views and ground-truth images. However, such fixed Gaussians generally render images from all views and lack the ability to adapt to specific viewpoints, as they do not incorporate target view information when predicting Gaussians. To address this, our UniqueSplat learns the view-conditioned information as a prior and incorporates this knowledge into network parameters, so that Gaussians are dynamically adjusted in accordance with different views. Specifically, we propose a two-branch view-conditioned hyperNetwork to simultaneously learn view-agnostic embeddings and view-specific knowledge, which not only explores the shareable knowledge from various views, but also adapts the model to specific views at test time. Extensive experiments on widely-used datasets including RealEstate10K, ACID and DTU demonstrate the superiority of UniqueSplat over the state-of-the-art methods. Moreover, UniqueSplat encouragingly outperforms existing methods in cross-dataset evaluation, showing its notable generalization ability.
Authors: Zhongjie Ba, Shengwang Xu, Peng Cheng, Jinyang Zou, Ting Yu, Zhibo Wang, Zhan Qin
Abstract: Embodied intelligence and world models require video understanding systems to go beyond recognizing objects and actions and develop an understanding of physical regularities. However, despite their strong performance on general video understanding tasks, current video-language models still struggle to reliably determine whether an observed event conforms to specific physical laws. Existing benchmarks primarily assess the physical quality of generated videos, providing limited support for systematically evaluating and improving the physical-law understanding of Video Large Language Models (VideoLLMs). To address this gap, we introduce PhyCheck, a video question answering dataset organized at two complementary levels of granularity. The coarse-grained subset asks models to determine whether the phenomenon shown in a video conforms to or violates physical laws, while the fine-grained subset further examines whether models can capture physical details responsible for the violation or compliance. We use these subsets as structured supervision to improve physical understanding. In addition, the dataset contains a diagnostic subset with external causal context that reveal hidden factors affecting physical plausibility, assessing whether models can recalibrate their judgments accordingly. Experiments with Fine-tune Qwen2.5-VL show that training with the proposed data substantially improves the understanding of physical-consistency, while evaluations in the diagnostic subset reveal that current models still have difficulty incorporating additional causal conditions into their decisions. These findings highlight the gap between recognizing surface-level inconsistencies and understanding underlying physical mechanisms, and provide a foundation for evaluating and improving physical understanding in Video-LLMs.
Authors: Xiaoke Yang, Haixu Song, Xiangyu Lu, Shao-Lun Huang, Yueqi Duan
Abstract: In this paper, we propose a characteristic-aware adaptive network named AdaForensics for deepfake detection. Most existing methods learn a fixed network to detect deepfakes based on carefully-designed network architectures. However, these methods employ the same deepfake detector for all the images despite of various facial characteristic, which fail to provide customized forgery detection for different individuals. To address this, our AdaForensics simultaneously learns characteristic-agnostic and characteristic-specific embeddings, where the detector dynamically adapts to varying faces with our designed hypernetwork on the fly. More specifically, our AdaForensics not only explores the shareable abstractions from various deepfake images, but also adapts the detector to the given characteristic at test time. To achieve this, we propose a two-branch HyperNetwork to learn an adaptive deepfake detector, which automatically adjusts the parameters based on characteristic of the input. Extensive experiments on widely-used datasets including FaceForensics, Celeb-DF and DFDC demonstrate our AdaForensics outperforms the state-of-the-art works.
Authors: Fanyu Wang, Longgao Zhang, Junyi Chen
Abstract: Existing rainfall simulation methods for autonomous driving remain limited in physical controllability and multi-view consistency. This paper presents GSRAIN, a high-/low-frequency rainfall synthesis method for 3D Gaussian Splatting (3DGS) driving scenes. GSRAIN constructs a high-frequency raindrop model from measured rainfall data and generates low-frequency rainy appearance using a geometry-aware single-step diffusion model. The two effects are then fused in a unified 3DGS scene, enabling rainfall-intensity control over the range of 0--13~mm/h. The proposed method achieves a Fr\'{e}chet Inception Distance (FID) of 149.09, outperforming CycleGAN-Turbo (155.71) and WeatherEdit (157.94). Object-detection and closed-loop driving experiments further show that the generated scenes expose scene-dependent performance changes of the evaluated algorithms under controllable rainfall. These results indicate that GSRAIN provides an effective approach for constructing physically controllable, repeatable, and closed-loop-compatible rainy-weather test scenes for autonomous driving.
Authors: Chongjian Wang, Junjie Gao
Abstract: Automatic sleep staging is a critical role in sleep disorder diagnosis, sleep quality assessment, and long-term health monitoring; however, existing approaches suffer poor performance on ambiguous and transition-related sleep stages, caused by inadequate modeling of fine-grained intra-epoch structures and complex cross-region spectral dependencies. Traditional epoch-level encoders commonly fail to extract subtle temporal microstructures and intra-epoch cross-region interactions, resulting in unsatisfactory recognition accuracy for hard categories such as the N1 stage. To tackle these drawbacks, we propose SwinSleepNet, a hierarchical context-aware dual-stream framework that separately optimizes intra-epoch representation learning and inter-epoch contextual modeling. Concretely, we characterize each sleep epoch from two complementary perspectives: raw time-domain EEG signal and its time-frequency transformation. The time-domain branch adopts convolutional encoders to capture fine waveform temporal details, and the time-frequency branch uses Swin Transformer to extract local spectro-temporal features, hierarchical multi-scale information and long-range spatial dependencies. The multi-branch extracted features are fused into integrated embeddings, which are optimized by a bidirectional context module to capture cross-epoch temporal dependencies for final sleep stage classification. Comprehensive experiments on Sleep-EDF-20, Sleep-EDF-78 and SHHS datasets verify that our method achieves competitive overall performance, and exhibits stronger robustness and stability on difficult N1 stages and transitional epochs. The results prove that optimized intra-epoch representation learning based on hierarchical architecture greatly benefits automatic sleep staging tasks.
Authors: Saurav Sharma, Lorenzo Arboit, Nabani Banik, Sarah Meuli, Julia Alekseenko, Jan Liechti, Franziska Heitzinger, Michela Orsi, Didier Mutter, Daniel Gero, Philipp C. Nett, Beat P. Muller, Joel L. Lavanchy, Nicolas Padoy
Abstract: Fine-grained understanding of surgical activity is essential for context-aware assistance in the operating room, including safety monitoring, adverse event identification, and skill assessment. Surgical action triplets, defined as tuples of the form
Authors: Fuzhen Jiang, Changyue Shi, Chuxiao Yang, Xinyuan Hu, Wenjie Ye, Minghao Chen
Abstract: Although image deraining has advanced substantially, existing methods mainly focus on 2D image restoration. As spatial intelligence applications such as embodied AI and autonomous driving continue to emerge, reconstructing clean 3D scenes from sparse rainy views in a feed-forward manner becomes increasingly important. Existing feed-forward 3D Gaussian Splatting (3DGS) methods often assume clean inputs and collapse under rainy conditions. To this end, we present \textbf{\textit{DerainSplat}}, a feed-forward framework that reconstructs clean 3D scenes from only a few rainy views. To support this task, we build a large-scale multi-view derain dataset through a four-stage synthesis pipeline that sequentially models overcast illumination, depth-dependent haze, rain streaks, and lens raindrops, producing privileged weather factors. We introduce a weather net that predicts the weather factors from rainy context and yields two support maps. Scene support modulates cross-view cost-volume matching, while radiance support drives depth-aligned appearance fusion to fill corrupted pixels. The derived geometry evidence further attenuates Gaussian opacity to reduce spurious structures. A rainy cycle consistency re-renders clean views using the predicted factors and aligns them with rainy inputs. Extensive experiments show that \textbf{\textit{DerainSplat}} outperforms existing methods on various datasets, including RealEstate10K, ACID, Mip-NeRF360, and real-world rainy scenes, with strong cross-dataset generalization.
Authors: Jiashuo Chen, Cheng Dai, Yanan Hu, Fanglin Bao
Abstract: Thermal imaging remains effective under adverse illumination, yet passive long-wave infrared (LWIR) measurements often lack fine texture. Existing thermal texture imaging approaches commonly rely on spectral sensing or registered auxiliary modalities, incurring substantial data throughput or vulnerability to cross-modal degradation. We introduce T$^2$exture, a sparsely perturbed thermal texture imaging framework that aims to reconstruct temporally dense thermal texture sequences from densely sampled passive frames and a few actively perturbed keyframes. We define thermal texture as the residual between a source-on observation and its corresponding source-off passive state. Under sparse LWIR illumination and rapid quasi-steady paired acquisition, this residual attenuates the passive-emission background and approximates a source-induced reflected response, exposing localized material- and geometry-dependent texture. T$^2$exture reconstructs a dense sequence of this source-conditioned response through two stages. Stage 1 estimates the unobserved source-off passive state at each active instant from neighboring passive frames to obtain reliable differential texture anchors. Stage 2 combines sparse anchors with passive structural context near each target time to reconstruct the dense sequence. On the simulated benchmark, T$^2$exture adds only 0.20M parameters to AMT-L while improving PSNR by 6.66 dB. Extensive evaluations on simulated and real acquisitions further show clearer texture recovery and stronger structural preservation than representative VFI baselines. These results establish T$^2$exture as a practical framework for thermal texture imaging under sparse active acquisition.
Authors: Cheng Li, Renjun Gao, Boyi Fu
Abstract: Traffic police gestures are safety-critical perception cues for autonomous driving. A deployable recognizer must infer commands causally from continuous full-frame video, remain stable around transitional arm motion, and avoid over-trusting corrupted pose measurements. This study presents RSC-GestureNet, a reliability-aware selective causal recognizer, for Chinese traffic police gestures. The model treats pose confidence as a first-class signal: unreliable joints are down weighted during graph reasoning, temporal evidence is aggregated causally, and calibrated predictions are selectively emitted through a reliability-aware inference rule. We further introduce CTPGesture-C, a reproducible feature-level corruption benchmark with seven pose/RGB degradation families, and an RGB-level diagnostic in which corrupted frames are reprocessed by MediaPipe before recognition. On the complete official CTPGesture v1 split (134,424 labeled frames and 33,451 causal windows), RSC-GestureNet achieves 93.33+-0.24% accuracy, 91.71+-0.27% macro-F1, 91.69+-0.29% online macro-F1, 98.80+-0.07% Early@10, 0.153+-0.013 s TTC, and the best robust macro-F1 among evaluated methods. Under the same split and causal protocol, it exceeds reproduced traffic-specific MD-GCN and HLP-GCN baselines by 3.23-4.11 macro-F1 points and 2.15-3.07 online-F1 points. These results, together with calibration, selective-risk, statistical, adaptive-branching, and image-level re-extraction analyses, indicate that explicit pose-reliability modeling improves early, stable, and robust traffic-command recognition.
Authors: Hantang Li, Qiang Zhu, Xiandong Meng, Debin Zhao, Xiaopeng Fan
Abstract: Sparse-view 3D Gaussian Splatting Super-resolution is highly challenging since the sparse and low-resolution (LR) inputs lack sufficient geometric and high-frequency information for accurate reconstruction. To achieve high-quality reconstruction, existing sparse-view super-resolution methods adhere to two-stage pipeline that performs LR Gaussian reconstruction and then high-resolution (HR) Gaussian refinement, which directly results in stage-wise Gaussian transfer and reconstruction error accumulation. To this end, we propose CLEAR, a Conflict-aware Learning via Evidence-guided Adaptive Routing, as the first unified single-stage framework for Sparse-view 3D Gaussian Splatting Super-resolution. Specifically, CLEAR performs joint the optimization of authentic LR observations and external HR priors within a unified Gaussian representation. To mitigate the gradient conflicts introduced by sparse supervision during training, we propose a Gaussian-wise conflict-aware optimization strategy that regards the LR gradient as a reliable anchor and applies evidence-conditioned soft correction only to severe HR conflicts. Moreover, to recover high-frequency details, we introduce an evidence-guided Patch-to-Gaussian routing mechanism which estimates patch reliability and detail demand, lifts them into Gaussian space, and selectively routes high-frequency gradients and densification. Finally, we employ shared Gaussian dropout and a detached mid-training anchoring to enhance the robustness of training framework. Extensive experiments on both synthetic and real-world $4\times$ super-resolution benchmarks demonstrate that CLEAR consistently achieves state-of-the-art rendering quality and superior geometric fidelity.
Authors: Gil Sasson, Zachary Levine, Smadar Shilo, Sarah Kohn, Guy Lutsker, Anastasia Godneva, Adam Gabet, David Krongauz, Adina Weinberger, Yann LeCun, Randall Balestriero, Eran Segal
Abstract: Whole-body dual-energy X-ray absorptiometry (DXA) scans are routinely acquired to measure bone density and regional body composition, leaving their spatial structure largely unused. Here, we show that self-supervised learning (SSL) can convert raw DXA images into representations of systemic health. We introduce LeDXA, a vision model based on a joint-embedding predictive architecture (JEPA) that learns by predicting latent representations rather than reconstructing pixels. Trained from scratch on 11,540 unlabeled Human Phenotype Project scans, LeDXA was evaluated internally and on 47,400 external UK Biobank (UKBB) scans. It improved cross-cohort prediction of prevalent diseases and biomarkers beyond scanner-derived DXA measurements and DINOv3, a state-of-the-art general-purpose model, despite approximately 150,000-fold fewer training images and nearly 40-fold fewer parameters. Over a median 4.3-year UKBB follow-up, LeDXA improved incident disease prediction over tabular DXA measures, with the largest gains for hip and knee arthrosis and type 2 diabetes. For hip arthrosis, 66% of incident cases occurred in the highest-risk quartile versus 41% for tabular measures. Its representations predicted chronological age externally (r = 0.88; mean absolute error = 2.90 years), and the biological-age gap tracked broader disease burden and a 45% higher mortality hazard in the oldest-appearing quartile. The gap also decreased in women after starting hormone-replacement therapy, suggesting it may be modifiable. Genome-wide associations recovered mostly known body-composition and bone-density loci, and LeDXA embeddings were more heritable than DINOv3's. These findings reveal prognostic information in DXA images that conventional readouts discard, learnable with relatively little data and modest compute.
Authors: Kaiyuan Pu, Tiantian Yang, Dan Zeng
Abstract: Visual AutoRegressive Modeling (VAR) has excelled in natural image generation via next-scale prediction, but its use on topology-structured data like human skeletons is still unexplored. VARPose is proposed to adaptively densify 2D sparse poses, thereby enriching the anatomical information available for 3D lifting models. Our core contributions are twofold. First, we introduce a Granularity-agnostic Pose Tokenizer (GPT), which employs a single hybrid codebook and a residual quantization strategy to encode poses of varying densities into a unified, multi-scale discrete representation. Our results demonstrate the strong generalizability of this representation. By decoupling the representation from the projection, we can successfully decode novel pose granularities using a frozen codebook with a retrained decoder. Second, we propose UniSkelar, a unified autoregressive model that treats "joint density" as "scale". UniSkelar learns to predict the token sequence for the next density level in a coarse-to-fine manner, conditioned on the sparsest pose. VARPose not only outperforms state-of-the-art methods and generalizes to unseen granularities, but also confers tangible performance gains on downstream tasks, such as 3D Pose Estimation and Human Mesh Recovery, through 2D pose densification. Our code and model are available at https://github.com/BRL-SYSU/VARPose.git.
Authors: Yan Huang, Guowei Wang, Xu Wang, Kangjun Liu, Xin Lin
Abstract: Vision-language models (VLMs) such as CLIP exhibit remarkable zero-shot capabilities, yet their performance frequently degrades sharply under unexpected test-time distribution shifts. While Test-Time Adaptation (TTA) offers a promising solution, continuously adapting VLMs over an unlabeled test stream presents fundamental challenges. Conventional top-1-centric updates often reinforce errors by corrupting the local semantic geometry among related classes, while iterative adaptation exacerbates progressive bias accumulation, ultimately driving the model toward mode collapse. To overcome these coupled vulnerabilities, we propose Local Margin Restoration (LMR), a lightweight, one-step TTA framework. At the sample level, our Protected Margin Restoration (PMR) objective recovers local semantic geometry by shielding plausible near-top candidates from external hard negatives. Concurrently, to combat stream-level degradation, we introduce a dual-stage stabilization mechanism, featuring an Adaptive Margin (AM) controller and Bias Correction (BC), to dynamically disrupt progressive bias accumulation and prevent mode collapse. Extensive experiments on CIFAR-C, ImageNet-C, and ImageNet variants demonstrate that LMR consistently outperforms state-of-the-art TTA baselines, proving exceptionally robust and efficient even in challenging low-batch test-time regimes. Our code is available at https://github.com/DennisHuangYan/LMR.
Authors: Yizheng Wu, Jiashen Hua, Bing Deng, Jieping Ye
Abstract: Agentic multimodal reasoning extends passive image understanding by allowing VLMs to actively acquire and refine visual evidence through visual tool interactions. Effective visual tool use requires three capabilities: grounding tool calls in visual context, composing tools across multiple steps, and adapting reasoning to tool-returned observations. However, existing approaches largely focus on grounding within fixed tool spaces and rigid invocation patterns, leaving composition and adaptation insufficiently addressed. We present VC-Tooler, which learns visual tool use as a compositional and adaptive capability. To this end, we first build a trajectory bank through a hierarchical synthesis pipeline covering three capability levels: single-tool grounding, multi-tool composition, and diverse tool contexts and interfaces. We then train the model in two stages: a supervised cold start that establishes these capabilities, followed by reinforcement learning that encourages accurate, efficient, and context-aware visual tool use. VC-Tooler achieves state-of-the-art performance among open-source models on both general-purpose and agentic benchmarks, including $95.8\%$ on V* and $35.3\%$ on VTC-Bench, and shows promising transfer under richer tool settings at inference time. Project page: https://w1zheng.github.io/VC-Tooler
Authors: Jidong Kuang, Hongsong Wang, Jie Gui
Abstract: Zero-shot skeleton-based action recognition (ZSAR) aims to recognize unseen action categories by aligning skeleton features with textual semantics. However, existing methods rely on text-derived prototypes that inherently lack geometric structure and physical constraints, resulting in a pronounced \textit{semantic-kinematic gap}. To bridge this gap, we propose \textbf{GenPrior}, the first framework to exploit generative priors from pre-trained Text-to-Motion (T2M) models for ZSAR. Specifically, we introduce Dispersion-Gated Feature Fusion, which distills kinematic prototypes and intra-class dispersion from generative motion sequences and employs a learned gating network to adaptively inject reliable structural cues into textual embeddings while suppressing synthetic artifacts. Furthermore, we propose Generative Prototype Refinement, which leverages these generation-enhanced prototypes as anchors to mine high-confidence unseen samples, calibrating class prototypes toward the true distribution and thereby unleashing strong performance gains. Extensive experiments on NTU-60, NTU-120, and PKU-MMD demonstrate that GenPrior achieves state-of-the-art performance under both zero-shot and generalized zero-shot settings. Code is available at https://github.com/jidongkuang/GenPrior.
Authors: Haozhe Luo, Ziyu Zhou, Shelley Zixin Shu, Mauricio Reyes
Abstract: Recent vision-language models for chest X-ray understanding are largely built on image-report alignment and therefore rely heavily on MIMIC-CXR as the dominant pretraining source. While effective at scale, this paradigm underexplores an important alternative source of supervision: a range of existing multi-label classification datasets, which provide cleaner and more explicit disease signals than free-text reports, and can offer broader pathology coverage when combined across sources. However, learning from such heterogeneous datasets is nontrivial, as differences in label ontologies, annotation protocols, acquisition pipelines, and report styles can cause models to entangle clinical semantics with dataset identity, leading to poor transfer despite increased scale. In this work, we revisit radiology VLM construction from the perspective of harmonized multi-source learning. We propose HarMoE, a dataset-aware mixture-of-experts framework that learns shared cross-dataset medical semantics while confining source-specific variation to lightweight residual experts in deeper decoder layers. To further exploit clean supervision from labeled datasets, we train in a unified disease vocabulary with masked multi-dataset supervision, enabling the model to leverage complementary annotations without introducing false negatives. Experiments on large-scale chest X-ray benchmarks show that HarMoE consistently improves zero-shot classification, out-of-distribution transfer, and grounding over strong baselines. Our results suggest that building robust radiology VLMs requires moving beyond single-source image-report alignment toward structured knowledge construction from heterogeneous datasets with cleaner supervision and broader coverage. Code and the 873k harmonized dataset will be released at https://github.com/Roypic/harmoe.
Authors: Jiangling Zhang, Shuxuan Gao, Zeyu Chen, Yichao Liu, Yu Zhou
Abstract: Rapidly evolving Generative AI enables sophisticated visual text manipulations that increasingly evade current forensic detectors. Existing discriminative models often overfit specific forgery patterns, limiting their generalization to unseen, open-set attacks. To address this challenge, we propose a generative detector that localizes tampering by estimating the local restoration cost required to align a query image with authentic visual-text statistics, rather than by learning forgery-specific decision boundaries. Specifically, we introduce Sparse-Constraint Rectified Flow (SC-RF), a detector-oriented adaptation of Flow Matching for spatially sparse anomaly localization. We further mitigate data scarcity via self-supervised Artifact Injection and preserve high-frequency forensic traces using a pixel-space Forensic-DiT. Extensive experiments on three benchmarks show that our method achieves state-of-the-art performance, surpassing the runner-up by 3.2 and 4.8 percentage points in F1 and IoU, respectively. In particular, the proposed detector demonstrates strong zero-shot performance on challenging unseen text editing patterns. We further provide an auxiliary stress-test analysis showing that local harmonization produced by our model can weaken the statistical cues relied upon by existing detectors, offering a complementary vulnerability-analysis perspective.
Authors: Haomin Peng, Yongkang Li, Zhaoxiang Liu, Xiaojie Jin, Shiguo Lian, Yunchao Wei, Xinggang Wang
Abstract: Open-vocabulary segmentation identifies and segments objects from arbitrary textual descriptions. SAM 3 supports noun-phrase-guided segmentation and achieves competitive open-vocabulary performance through exhaustive vocabulary traversal, yet suffers from prohibitive computational overhead as target categories scale. In this paper, we propose an Efficient Open-Vocabulary segmentation framework with SAM 3 (EOVSAM), which adapts SAM 3 for single-pass prediction. EOVSAM removes prompt conditioning to turn SAM 3 into an efficient mask generator and introduces a new Attentional Aggregation strategy to optimize open-vocabulary classification end-to-end. This formulation avoids the multi-stage pipelines and post-processing heuristics commonly used by existing methods, while mitigating the closed-set collapse that can arise when classification is optimized directly. EOVSAM consistently improves segmentation accuracy over vanilla SAM 3 on all evaluated datasets and accelerates inference by up to 338$\times$. Furthermore, EOVSAM maintains high accuracy at lower resolutions while achieving even more remarkable inference speeds. Experiments on standard semantic and panoptic segmentation benchmarks show that EOVSAM combines competitive or state-of-the-art accuracy with a substantial speed advantage over existing open-vocabulary segmentation models. Code and models are available at https://github.com/hustvl/EOVSAM.
Authors: Aoru Xue (ShanghaiTech University, Shanghai, China), Yujing Sun (Digital Trust Centre, Nanyang Technological University, Singapore), Yiming Ren (ShanghaiTech University, Shanghai, China, Digital Trust Centre, Nanyang Technological University, Singapore), Kwok-Yan Lam (Digital Trust Centre, Nanyang Technological University, Singapore), Mao Ye (EABOT.AI, China), Yuexin Ma (ShanghaiTech University, Shanghai, China)
Abstract: We propose Sen-Cap, a Sensor-Flexible and Noise-Resilient 3D human motion Capture framework that integrates multi-modal data from LiDAR and camera. While multi-modal sensors provide richer information than single-modal sensors, existing approaches still suffer from two core challenges. First, multi-modal alignment/matching across arbitrarily deployed sensors is typically handled by explicit calibration, which propagates errors under changing viewpoints and in turn constrains deployment to fixed, highly overlapped layouts. Second, prior methods degrade under severe noise or partial sensor failures, which are common in real-world environments. To address these challenges, Sen-Cap introduces a Unified Across-Sensor Motion Estimator that reconstructs local pose and shape in a human-centric space without calibrations between sensors, supporting a flexible number of sensors, as well as a Noise-Resistant Trajectory Tracker that maintains robustness under severe point cloud noise through iterative refinement. These sensor-flexible and noise-resilient features make Sen-Cap more practical in real-world deployment. Notably, operating in real time, Sen-Cap achieves state-of-the-art performance on major metrics on Human-M3 and FreeMotion, as well as strong cross-domain performance on LiDARHuman26M and RELI11D. This combination of flexibility and robustness opens new opportunities for motion capture in real-world scenarios, e.g. sports analytics, field robotics, and large-scale immersive environments.
Authors: Annajoyce Mariani, Kira Maag, Hanno Gottschalk
Abstract: Realistic and diverse trajectory generation is central to enabling higher levels of vehicle automation. While rule-based and classical learning-based methods may struggle to capture the complexity of traffic behavior, generative models have already demonstrated in other fields that they can handle a comparable level of complexity. In this paper, we build upon previous work on generative adversarial network (GAN)-based semantic bird's-eye-view traffic generation and extend the proposed framework in several key aspects. We improve the semantic representation, replace the trajectory extraction procedure with a graph-based association method, and systematically investigate increasingly larger fields of view. In addition, we introduce a quantitative evaluation framework to assess hallucinations and object permanence in generated videos. Our experiments demonstrate that the framework generalizes to larger and more complex traffic scenes while maintaining statistically realistic trajectories and coherent spatial relationships between traffic participants. Within 150GPU hours of training and with inference times below 20ms for scenes of up to 20s, our results demonstrate that video-based GANs remain an efficient and scalable approach for realistic trajectory generation, even in substantially larger traffic scenes, making them well suited for downstream tasks such as prediction, planning, and simulation in automated driving.
Authors: Shengkai Hu, Jie Shao, Jiaqi Ma, Xu Zhang, Keying Wu, Qilu Zhu, Beihang Song, Jun Wan
Abstract: ANN-based All-in-One image restoration (AiOIR) unifies diverse degradation handling but incurs high computational costs, limiting its real-time deployment. While Spiking Neural Networks (SNNs) offer a low-power alternative, applying them to static images remains challenging. This difficulty arises because explicit event signals are absent, and degradation cues are heavily entangled with scene structures, hindering the learning of reliable restoration-oriented spike events. To address these issues, we propose SpikeRestormer, an energy-efficient SNN for AiOIR that performs event reasoning over internally generated spike cues. Specifically, we propose a degradation-event perception process to extract spike-based degradation events through Subtractive Degradation Event Attention (SDEA). Moreover, we introduce Hierarchical Bayesian Skip Masking (HBSM) and Additive Restoration Event Attention (AREA) processes for event-reliability inference and restoration-event construction, respectively. By integrating these complementary processes, SpikeRestormer formulates restoration as a unified process of degradation-event perception, degradation-event reliability inference, and restoration-event construction, liberating the potential of SNNs for energy-efficient AiOIR. Extensive experiments show that SpikeRestormer delivers competitive performance against ANN-based methods and establishes new state-of-the-art results among SNN-based methods with significantly lower energy consumption.
Authors: Dichang Zhang, Jiaqi Deng, Yixuan Shao, Yuanpeng Liu, Jiali Cui, Zhiqiang Lao, Heather Yu, Liang Peng, Simon Birrer, Dimitris Samaras
Abstract: Existing astronomy foundation models provide strong galaxy representations, but adapting them to new survey conditions and survey-specific morphology recognition tasks still requires substantial human supervision. We show that VLM-based VQA systems contain meaningful visual-semantic priors that can serve as weak supervision for downstream morphology classifiers and improve morphology classification under limited human-label budgets. We first introduce a survey-oriented VQA benchmark spanning two representative imaging regimes and evaluate state-of-the-art VLMs on galaxy morphology questions. The results show that these models capture useful morphology signals and informative uncertainty, but are not sufficiently reliable to replace human annotators. Motivated by this finding, we use a general-purpose VLM as a morphology teacher for Zoobot, an astronomy foundation model pretrained on large-scale Galaxy Zoo annotations. Across two survey domains and multiple annotation budgets, the VLM teacher consistently improves Zoobot's downstream morphology classification. These results demonstrate that a general-purpose VLM provides knowledge complementary to an astronomy foundation model and can teach it to better recognize galaxy morphology under limited human supervision. The resulting pipeline is designed for label-efficient adaptation to forthcoming large-scale surveys, including the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) and the Nancy Grace Roman Space Telescope. The benchmark and code are publicly available at https://github.com/fw-ic/VLM-morphology-teacher.
Authors: Roua Rouatbi, Juan-Esteban Suarez Cardona, Ivo F. Sbalzarini
Abstract: We introduce a mathematical framework for shape comparison based on mapping functions from the shape domain to a common reference domain. This Push-Forward Transform enables invariant and robust comparison of shapes, preserving intrinsic geometric information. Quantitatively comparing shapes and their temporal evolution is a fundamental challenge in image analysis. Meaningful shape comparison requires representations that are invariant to transformations that do not alter shape itself, such as translation, rotation, reflection, re-parametrization, and uniform scaling, while remaining sensitive to intrinsic geometric variation. Existing approaches often rely on sensitive parameterizations, landmark correspondence, or learned representations that are difficult to interpret and reproduce. We show that the Push-Forward Transform (PF-T) applied to Signed Distance Functions (SDFs) yields a continuous representation that captures both boundary and interior geometry. We derive an interpretable morphometric that quantifies shape similarity and reveals features such as skeletal topology and rotational symmetries. The push-forward transform applies consistently to two- and three-dimensional shapes, extends to time-evolving geometries, and supports the joint analysis of shape and additional scalar fields defined over shapes, such as intensity or molecular signals. We present the mathematical formulation, describe an efficient algorithm, and benchmark the approach on 2D, 3D, and temporal data sets.
Authors: Filippo D'Addeo, Lorenzo Cipelli, Adriano Cardace, Emanuele Ghelfi, Andrea Zinelli, Massimo Bertozzi
Abstract: We present CalibBEV, a novel Bird's Eye View (BEV) alignment approach for LiDAR-camera calibration. Our method unifies LiDAR and camera data into a shared 3D spatial representation, enabling accurate and robust cross-modal calibration. CalibBEV extracts sensor-wise BEV features from each modality using domain-specific architectures and estimates the calibration matrix through a two-step alignment process. First, we perform an implicit alignment by regressing a coarse calibration matrix directly from the BEV features. To ease this alignment, we enforce semantic consistency between BEV representations across modalities using a contrastive loss inspired by CLIP, guiding both networks toward a unified feature space. In the second step, we leverage our BEV formulation to explicitly align the features of one modality with the other, refining the initial coarse estimate into a final, more accurate calibration matrix. CalibBEV significantly outperforms prior point-to-pixel matching methods, achieving state-of-the-art calibration accuracy. On the KITTI and nuScenes benchmarks, our method reduces the Relative Rotation Error (RRE) by 51% and 68%, and the Relative Translation Error (RTE) by 80% and 91%, respectively, compared to previous methods.
Authors: Gaetano Chiriaco, Luca Barco, Andrea Bragagnolo, Claudio Rossi, Edoardo Arnaudo
Abstract: Geospatial foundation models aim to learn representations that transfer across regions and sensors, yet evaluating them on specific tasks requires large, high-quality, multi-modal benchmarks that measure how well such models extract value from data. Concerning flood mapping, existing datasets rarely combine bi-temporal SAR and co-registered optical imagery at scale, leaving the value of foundation models for this downstream task largely untested. We introduce GEOID-Flood, a large-scale multi-modal flood segmentation benchmark, derived from Copernicus Emergency Management Service activations, spanning 219 events across 65 countries over ten years. The dataset provides more than 14,000 tiles with co-registered pre- and post-event Sentinel-1, in GRD and RTC format, pre-event Sentinel-2 composite, and DEM, including manually validated labels that separate background from permanent water and flooded water. Using this benchmark, we evaluate foundation models against conventional encoders across single-image, multi-temporal, and multi-modal protocols. We report three main findings: foundation models offer a consistent but modest advantage; optical-SAR fusion with finetuning best resolves transient flooding; and models trained on GEOID-Flood transfer to unseen events better than those trained on existing datasets. Dataset and code available at https://github.com/links-ads/geoid-flood.
Authors: Ang Li, Menghui Jiang, Xiaobin Guan, Dong Chu, Huanfeng Shen
Abstract: Accurate and efficient reconstruction of cloud-contaminated and noise-corrupted NDVI time series remains a challenge in remote sensing. Deep learning provides a promising solution for modeling complex spatiotemporal dependencies; however, its application is often limited by the difficulty of obtaining paired clear-sky and degraded NDVI data for identical spatiotemporal locations. To address this issue, we propose GloSSR, a Global-scale Self-supervised Spatiotemporal framework for NDVI Reconstruction. The framework constructs supervisory signals by artificially degrading relatively clean NDVI observations with realistic cloud contamination patterns, producing self-supervised training pairs that closely mimic real-world degradation. It further introduces an end-to-end spatiotemporal learning network that jointly captures long-range temporal dependencies and short-term spatiotemporal correlation through a bidirectional Transformer with a ConvLSTM architecture. A temporal-channel attention-based reconstruction module is incorporated to enhance informative features, while a spatiotemporal prior constraint is designed to preserve both fine-scale structures and long-term phenological trends during optimization. Extensive evaluations on MODIS NDVI data demonstrate the effectiveness of the proposed framework across both artificial and real-world scenarios. In artificial degraded-pixel reconstruction experiments, GloSSR consistently outperforms the comparison methods. Time-series analyses based on real observations further demonstrate that the proposed framework can accurately characterize vegetation dynamics and capture the key phenological states. Long-term vegetation trend analysis and the transferability analysis to AVHRR data validate the scalability of the framework and illustrate its broad applicability for large-scale environmental monitoring.
Authors: Sina Wendrich, Lukas F\"orner, Zoe Reinke, Kartikay Tehlan, Ansgar Berlis, Michael Fr\"uhwald, Matthias Wagner, Thomas Wendler
Abstract: Longitudinal multiparametric MRI is central to follow-up imaging in oncology, yet real-world clinical data are characterised by missing sequences, heterogeneous acquisition protocols, and varying spatial resolutions across time points. We propose a patient-specific conditional implicit neural representation (INR) that models multimodal longitudinal MRI as a continuous function of world coordinates, time, and modality conditioning. The model is trained with stochastic modality dropout to handle incomplete data, and its continuous coordinate-space formulation enables both spatial and temporal interpolation without resampling to a fixed voxel grid. A self-consistency-based confidence estimator is derived from cross-modal reconstruction performance at inference time. We evaluate the framework on longitudinal MRI from paediatric brain tumour patients, demonstrating statistically significant improvements over linear interpolation for T1CE and FLAIR (p < 0.05), with mean MS-SSIM of 0.95 $\pm$ 0.02 for T1CE. Predicted confidence correlates strongly with true reconstruction quality (Pearson r up to 0.996), suggesting reliable deployment potential in heterogeneous clinical settings.
Authors: Mohammad Mahdi Dehshibi, David Masip
Abstract: The same body posture can convey entirely different emotions depending on its surrounding context, yet most methods for recognising bodily emotions treat scene and object cues as auxiliary feature augmentations rather than as structured priors over the plausibility of emotions. We introduce the Context-Aware Mixture of Domain Experts (CA-MoDE) for bodily emotion recognition. CA-MoDE incorporates dedicated scene and object experts to generate soft distributions over emotion categories conditioned on their respective domains. These domain-conditioned soft predictions serve as structured contextual priors that modulate the body expert's predictions at the distributional level rather than at the feature level. To fuse these multi-domain signals, we propose a task-tailored max-endorsement gating strategy that selects the strongest contextual signal across experts for each emotion dimension. Our gating strategy mitigates the signal dilution that typically occurs when conflicting or uninformative context distributions are averaged. CA-MoDE achieves an Emotion Recognition Score of 0.3269 on the Body Language Database. By outperforming existing temporal models using only single still images, our framework demonstrates that explicitly modelling structured spatial context can serve as a complementary discriminative proxy for the behavioural dynamics typically captured by video.
Authors: Runci Bai, Yucheng Xin, Pu Wang, Yongcong Wang, Chen Wu, Dianjie Lu, Guijuan Zhang, Pengwen Dai, Guangwei Gao, Siyuan Yao, Zhuoran Zheng
Abstract: Old photographs often suffer from multiple coupled degradations, including scratches, cracks, fading, blur, noise, and missing regions, severely degrading both visual quality and semantic content. We propose Loop-Mamba, a lightweight loop-based state-space framework that formulates old photo restoration as progressive state evolution, where a persis- tent restoration state is continuously propagated and refined through iterative computation. Specifically, we introduce a Semantic-Guided Degradation Estimator (SGDE) to explicitly model heterogeneous degradations by jointly predicting local degradation maps and global degradation scores, providing degradation-aware guidance for state evolution. We further develop a Shared Structural Memory Mamba (S$^2$M- Mamba), which maintains a persistent restoration state across iterations, enabling persistent state evolution through shared structural memory for robust long-range structural reconstruction. Benefiting from first-order state recursion, Loop-Mamba propagates latent restoration states through recurrent tran- sitions instead of repeatedly stacking deep feature transformations, thereby alleviating gradient dilution while avoiding the computational overhead inherent in iterative CNN- and Transformer-based restoration frameworks. A lightweight multi-directional scanning strategy further enhances direc- tional information aggregation and preserves structural continuity. To better evaluate restoration quality, we introduce the task-oriented Old Photo Damage Recovery Score (ODRS), which jointly measures degradation recovery and structural reconstruction fidelity. Experimental results on the public SynOld benchmark demonstrate that Loop-Mamba consistently outperforms previous state-of-the-art methods across both conventional restoration metrics and the proposed ODRS.
Authors: Sitong Gong, Caixin Kang, Tianyu Yan, Guo Chen, Bo Zheng, Kaipeng Zhang, Yunzhi Zhuge, Xiang Ruan, Huchuan Lu, Yifei Huang
Abstract: A wearable assistant should both answer questions about its visual history and recognize when that history is useful to the present situation. Existing video-memory systems primarily support question-conditioned recall, whereas proactive assistants typically use separate memory and control mechanisms. We introduce GROVE, a training-free framework that supports both behaviors with one memory grown causally from a continuous video stream. GROVE retains fine-grained perceptual evidence and incrementally consolidates it into time-stamped moments, coherent episodes, and recurring cross-day patterns. Each stratum is paired with a scale-native retrieval skill for locating an observation, replaying an activity, or traversing long-range regularities. Reactive QA and proactive assistance share this memory and access interface, differing in whether retrieval is initiated by a user query or the current situation. Across multiple benchmarks including the challenging MM-lifelong and EgoServe, GROVE achieves the best results among the compared methods. Controlled ablations show that the temporal strata and their access skills are complementary, with patterns providing the largest benefit when evidence spans multiple days. Code will be available at https://github.com/SitongGong/GROVE.
Authors: Sathiyamohan Nishankar, Nethmi Pathirana, Pubudu Sanjeewani, Asanka Perera, Selvarajah Thuseethan
Abstract: Most evidence on the effectiveness of explainable artificial intelligence (XAI) attribution methods has been established on convolutional neural networks (CNNs), with limited investigation into whether these conclusions generalize to the diverse Vision Transformer (ViT) architectures that now dominate computer vision. This paper presents a controlled benchmark that evaluates attribution quality across five dimensions: faithfulness, localization, robustness, complexity, and computational cost. A standardized framework assesses 13 attribution methods from four algorithmic families on eight representative backbones spanning CNNs, isotropic ViTs, hierarchical transformers, hybrid architectures, and linear-attention transformers. The results show that attribution performance is strongly architecture-dependent and that rankings established on CNNs do not reliably transfer to transformer-based models. CAM-based methods achieve the highest scores under the conventional bounding-box localization metric on CNNs and most ViTs but perform poorly on linear-attention architectures. Pixel-level dense-mask evaluation further reveals that these gains largely reflect metric saturation rather than accurate localization. CAM-based methods also exhibit limited robustness on global-attention transformers, whereas attention rollout provides consistently stable explanations with poor localization. Furthermore, faithfulness correlation offers limited discrimination between attribution methods, highlighting the limitations of single-metric evaluation. These findings challenge prevailing conclusions on attribution performance and demonstrate the need for architecture-aware, multi-dimensional evaluation. The open-source code for the evaluation framework and benchmark results is available at https://github.com/Nishan-Charlie/VIT_XAI_Bench.
Authors: Shi Chen, Jie Zhang, Yicong Zhou
Abstract: Hyperspectral image super-resolution aims to reconstruct high-resolution imagery while preserving dense spectral information. Recently, Mamba-based models have shown promising potential for this task by capturing long-range dependencies with linear computational complexity. Nevertheless, their causal sequence modeling requires two-dimensional hyperspectral features to be unfolded along predefined scanning orders, which disrupts spatial adjacency and restricts the effective propagation of contextual information. Moreover, state-space parameterization of existing models is predominantly derived from generic learned representations, without explicit alignment with the intrinsic characteristics of the hyperspectral image. To address this issue, we propose an Unmixing-derived Spectral and Structural Prompting Mamba framework, termed USP-Mamba, which adapts Mamba state evolution through composition-aware spectral priors and image-dependent structural prompts. Specifically, an unmixing-informed spectral prompt captures the global material composition of the input image and provides persistent conditioning throughout reconstruction. Injected into the Mamba sequence and progressively adapted across layers, it steers state evolution toward composition-consistent reconstruction. We introduce feature-level structural prompts comprising spatial and frequency components to provide image-dependent local guidance. The spatial prompt promotes structure-sensitive state encoding for local detail preservation, while the frequency prompt enables region-adaptive transitions between homogeneous regions and high-frequency details. Finally, complementary Hilbert and Semantic-Guided Neighboring scans preserve spatial continuity and strengthen non-local semantic dependency modeling. Extensive experiments on different datasets demonstrate that the proposed method consistently outperforms representative approaches.
Authors: Scott McAvoy, Jonathan Klingspon, George Bent, Dave Pfaff, Aviral Agarwal, Maurizio Seracini, Falko Kuester
Abstract: The Loggia dei Lanzi in the Piazza della Signoria is one of Florence's most prominent structures visited by millions every year. Its construction history spans multiple centuries of modification. This paper presents the results of a thermal imaging campaign conducted in December 2025, using a FLIR T1020 HD camera, revealing hidden architectural features including walled-up openings and material transitions beneath the plaster surface. The favorable winter ambient conditions provided a feature-rich benchmark upon which to compare the results of enhancement algorithms and artificial intelligence models. We evaluate the application of AI-based image enhancement to thermal heritage documentation through a comparison of three tiers of image resolution in a photogrammetric Structure-from-Motion (SfM) pipeline: native resolution, FLIR's hardware-based pixel-shifted super-resolution (UltraMax), and state of the art AI-upscaled imagery models. We quantify the effect of each resolution tier on feature detection and tie-point generation, assessing whether the additional detail produced by super-resolution, whether hardware or AI-derived, translates into meaningfully denser and more accurate 3D thermal models. Our results contribute to the emerging intersection of artificial intelligence and heritage thermography by providing a direct comparison of hardware microscanning and AI super-resolution within a thermal photogrammetric workflow for cultural heritage. All datasets are made publicly available and accessible within an interactive 3D archival framework, and integrated into a custom citywide extended reality overlay application.
Authors: Jiaming Chen, Guoan Xu, Aoshen Huang, Haozhuo Zhang, Yang Li, Wei Pan
Abstract: Forecasting future states from video sequences is a critical challenge for autonomous robotic systems and a fundamental objective of world modeling. Prior generative methods operating at the pixel level inevitably overemphasize task-irrelevant details, leading to prohibitive computational overhead. While latent-based approaches attempt to mitigate this by predicting features directly, the persistent reliance on heavy decoders for state-to-task mapping remains a computational bottleneck. In this work, we propose Decoder-Free Feature Forecasting (DF$^3$), a novel framework that models world evolution entirely within the latent space and directly derives task outputs, completely eliminating the need for a decoder. Specifically, DF$^3$ injects learnable spatial queries into the terminal blocks of a frozen vision foundation model to extract future state representations directly. By employing a lightweight, unified Motion-Aware Context Fusion (MACF) mechanism that seamlessly integrates coarse flow warping with fine-grained latent cross-correlation, these queries interact with historical token representations to explicitly align and forecast the feature of the next frame. Subsequently, a specialized set of task queries probes these forecasted features for the downstream task. Extensive experiments on public benchmarks and zero-shot deployment in a robotic simulator demonstrate that DF$^3$ achieves performance comparable to state-of-the-art methods while offering superior efficiency and flexibility for integrated perception and control.
Authors: Monan Sun, Bangzhen Liu, Huaidong Zhang, Shengfeng He
Abstract: Autoregressive models have emerged as an effective paradigm for point cloud generation. However, most existing approaches rely on heuristic tokenization strategies, such as spatial sorting or stochastic downsampling, which often disrupt intrinsic point cloud topology and weaken the structural coherence of the generated shapes. In this paper, we present PointRSP, an autoregressive framework that reformulates point cloud generation as a topology-preserving tessellation process via recursive spectral partitioning. Instead of constructing token sequences heuristically, we introduce a topology-aware partitioning autoencoder that decomposes an unstructured point cloud into a non-balanced binary tree through a hybrid recursive spectral partitioning strategy. This hierarchical representation provides a deterministic geometric blueprint that preserves topological relationships while capturing multiscale structural dependencies within a quantized latent space. To synthesize shapes in this space, we propose a dual-stream cascaded generator that jointly models structural evolution and feature synthesis. In addition, we design a geometry-calibrated positional encoding mechanism that anchors latent embeddings using multi-scale structural centers, which stabilizes cascaded generation during the early stages of structural formation. Extensive experiments show that PointRSP achieves state-of-the-art performance in generation quality and diversity, demonstrating strong generalization across complex 3D topologies.
Authors: Jiawei Wang, Hao Yu, Yongzhen Hu, Xinyi Yang, Tao Ni, Xin Zhan, Junbo Chen, Xiaowei Zhou, Ruizhen Hu, Sida Peng
Abstract: Single-image feed-forward 3D Gaussian Splatting (3DGS) aims to directly generate a renderable 3D scene representation from one input image, avoiding the cost of multi-view capture and per-scene optimization. However, existing methods are often constrained by a pixel-aligned representation, where Gaussians are predicted from fixed image-grid locations. Such pixel-aligned primitives can produce promising nearby-view renderings, but they remain weakly coupled to underlying scene surfaces and struggle to preserve coherent structures under large viewpoint shifts. We present InfiniSplat, a feed-forward single-image 3DGS framework that moves from a pixel-aligned representation toward a surface-aligned representation. InfiniSplat constructs this representation by first using geometry-guided sampling to place 2D supports according to depth-induced local surface structure, and then applying a query-conditioned implicit decoder to predict Gaussian attributes from the image features queried at these supports. By grounding support locations in geometry while decoupling Gaussian prediction from fixed pixel centers, InfiniSplat produces Gaussian layouts that better follow scene surfaces and reduce scattered primitives caused by grid discretization. Across multiple cross-dataset NVS evaluations, InfiniSplat achieves state-of-the-art performance compared with single-image feed-forward baselines, and demonstrates zero-shot generalization from Hypersim indoor synthetic training to complex open-world scenes. Project page: https://zju3dv.github.io/InfiniSplat.
Authors: Yonatan Ben Avraham, Baruch Binyaminov, Yehudit Aperstein
Abstract: Rip currents are recurrent coastal natural hazards that threaten beachgoers and create operational challenges for lifeguards and coastal managers. Reliable monitoring from standard RGB (red-green-blue) imagery acquired by unmanned aerial vehicles (UAVs) remains difficult because hazardous channels often appear as subtle gaps in breaking waves, foam texture, or sediment patterns, and these signatures are affected by illumination, sea state, and environmental noise. This study presents a physically informed coastal environmental monitoring workflow for detecting visually expressed rip-current indicators that integrates wavelet-derived spatial-frequency texture features with deep learning. We evaluate multiple strategies for incorporating Discrete Wavelet Transform features into convolutional architectures, from computationally efficient channel replacement to dual-stream fusion with attention mechanisms. Performance is assessed against a standard RGB baseline using a task specific convolutional neural network for image-level presence classification and a YOLOv8 model for object-level localization. Under the evaluated dataset conditions, integrating wavelet derived texture features improves performance over RGB-only models. The dual-stream architecture achieves the strongest classification performance, exceeding 95% accuracy with high recall, while channel replacement is most effective for YOLOv8 object detection, reaching 94% mAP@50 for localization. Explainable artificial intelligence analyses provide qualitative evidence that the models attend to visually plausible wave-gap regions associated with rip currents. These results suggest that under the conditions of the evaluated dataset, physically informed wavelet integration may support UAV-based decision-support tools for interpretable beach-safety risk mitigation.
Authors: Ambarish Govindarajulu Kaliamurthi (San Jose State University), Kaikai Liu (San Jose State University)
Abstract: Deploying vision-language models (VLMs) for safety-critical spatial reasoning on resource-constrained autonomous driving platforms requires both compact model size and reliable metric grounding. We present MoRAL (Multimodal Reasoning for Autonomous Language Models), a two-stage fine-tuning pipeline that teaches Cosmos-Reason2-2B to first read a physics-encoded Bird's Eye View (BEV) representation and then reason over it for driving decisions. The BEV image encodes LiDAR metric distance as color bands, object class as cluster morphology, and radar Doppler velocity as directional wedge overlays, externalizing spatial perception into the input image so that no learned 3D backbone is required at inference. Stage 1 fine-tunes the vision encoder on 60,000 grounding records; zero-shot baselines produce no parseable BEV outputs, confirming the vocabulary requires explicit training. Stage 2 fine-tunes the full model (52M parameters, 2.4% of total) on 57,696 chain-of-thought records generated by Cosmos-Reason2-8B as teacher, spanning eight driving question types. On 2,304 held-out nuScenes frames evaluated by Gemma 4 (31B) calibrated against human review, MoRAL wins seven of eight question types over a zero-shot 8B baseline despite using four times fewer parameters, with the largest margins on question types requiring structured multi-step physics reasoning. Emergency braking recall improves from 10.8% to 47.8%, output degeneration falls from 94.1% to 20.8%, and the full pipeline fits a consumer 8 GB GPU at 42 tok/s without quantization. These results establish a reproducible foundation for compact, physics-grounded VLM reasoning on mobile edge platforms.
Authors: Thanh Duc Pham, Anh Nguyen, Duong Duc Hieu, Minh-Tan Pham
Abstract: Interactive segmentation reduces the prohibitive cost of pixel-level annotation by allowing users to delineate objects with a few clicks. However, applying this paradigm directly to remote sensing imagery is non-trivial: ultra-high resolutions, small object sizes, and sparse spatial distributions all degrade segmentation quality. Recent work has addressed the resolution barrier and achieved competitive results in interactive segmentation for remote sensing (ISRS). However, they treat all instances of a class within an image as a single objective target. Consequently, interactions spent on one object contribute nothing to its same-class neighbours, and satisfactory masks may demand up to 40 clicks per image, hindering the practicality of these frameworks. We observe that remote sensing scenes exhibit markedly strong inter-object correlation, meaning a single clicked object is highly informative about the rest of its category. Building on this, we propose ISRS-DETR, a detection-guided interactive segmentation framework that injects object-level evidence into both training and inference. Our ISRS-DETR employs an RF-DETR decoder with the interactive segmentation backbone to localise co-occurring same-class objects, and introduces a Dynamic Top-K Click Selection strategy that retains only reliable proposals and converts each into a simulated click, so one user interaction propagates across an entire class. Experiments on three standard remote sensing benchmarks show that ISRS-DETR achieves state-of-the-art accuracy while substantially reducing Number of Clicks per Image (NoC-I). All codes and data splits will be released for reproducibility upon acceptance.
Authors: Mohamed ElBassat, Seifeldin Elkerdany, Mohamed ElBialy, Gamal Abouelhamd, Jana Ghoneim, Assem Elkady, Mohamed Elboraay, Nelly Semenova
Abstract: AnimalCLEF26 addresses discovery-oriented animal re-identification, where systems must both attach query images to known individuals and discover unseen individuals by clustering them correctly. We present a similarity-to-clustering pipeline for this setting across Eurasian lynx, fire salamander, loggerhead sea turtle, and Texas horned lizard images. The method first isolates the target specimen using segmentation and then applies lightweight species-specific preprocessing for lynx, sea turtle, and salamander images to enhance identity-relevant visual cues, while Texas horned lizard images are used after segmentation only. Pairwise similarities are then estimated with WildFusion by calibrating and combining a MiewID global descriptor with two local matching branches, ALIKED + LightGlue and DISK + LightGlue. The resulting query-query similarities are refined and converted into identity clusters using graph-based clustering, while query-database similarities are used to attach confident samples to known identities. We evaluate training-free and fine-tuned MiewID variants, including Dynamic ArcFace and SphereFace2-Focal adaptations, and combine them in the final ensemble. Our selected ensemble substantially improves on the WildFusion baseline, achieving the best public ARI of 0.72124 and a private ARI of 0.70393, while a simpler preprocessing-before-calibration variant achieves the best private ARI of 0.71087. These results indicate that calibrated global-local fusion with species-aware preprocessing choices is effective for open-set wildlife re-identification under challenging field conditions and visual variation. The implementation code is available on GitHub.
Authors: Vishwajeet Shivaji Hogale, Anjali Pai, Nitya Ravi
Abstract: Vision-language models (VLMs) are increasingly deployed as reasoning agents in real-world visual assessment pipelines, yet their spatial grounding remains unreliable for fine-grained, visually ambiguous targets. We study this gap in the context of automated vehicle damage assessment, where fine-grained defects such as scratches and hairline cracks occupy few pixels, produce weak gradient signal, and are easily confused with reflections and surface texture. We show that a state-of-the-art VLM (Qwen-VL) achieves strong semantic classification accuracy (87.3%) on this task but is systematically ungrounded at the spatial level: it hallucinates damage in reflective regions, misses elongated scratches entirely, and produces spatially inconsistent outputs when prompted for localization. We propose TinyDamage, a hybrid architecture that delegates spatial grounding to a dedicated multi-task segmentation model while reserving the VLM for semantic reasoning and report generation. On the segmentation side, we find that the choice of loss function has an outsized and underexplored effect on tiny-object grounding: focal loss, widely used for class imbalance, collapses tiny-damage detection to zero, while a supervised contrastive objective measurably improves damage/background separability. We integrate the segmentation model into a 7-node LangGraph agent pipeline that grounds every VLM generation step in the segmentation output, and show that this grounding reduces the report hallucination rate from 92% (text-only) and 78% (image-only) to 31% in a controlled evaluation on 100 human-verified reports. We introduce DET_l, a permissive per-category detection metric for evaluating tiny-object grounding under class imbalance, and report latency and reliability characteristics of the deployed pipeline.
Authors: Jiayu Gu, Yiwei Wang, Jie Zhang, Guojun Cao, Keshen Lyu, Song Zhou, Yimeng Chen, Haorui Wang, Qingmin Feng, Shenchao Shi, Huan Zhao, Wenbin Chen, Caihua Xiong, Chidan Wan, Jing Samantha Pan, Xiong Cai, Han Ding
Abstract: Computational attention models could help surgeons manage the visual demands of laparoscopy, but they require dense spatial labels that are difficult to obtain because surgical intent is highly specialized and tacit. Here, we introduce DiffeoAfford, an action-grounded tissue affordance framework that retrospectively derives visual attention supervision from completed surgical procedures. By combining diffeomorphism-constrained tissue tracking with instrument trajectory analysis, DiffeoAfford generates affordance hotspot labels without manual per-frame annotation. A real-time prediction model trained on these labels anticipates relevant surgical regions and enables AffordView, an assistive auto-framing system for laparoscopic visualization. The proposed framework aligns with expert annotations and intraoperative surgeon gaze, and reduces surgeon cognitive workload during real-world evaluations using subjective, physiological, and behavioral measures.
Authors: Jiayu Chen, Xiaoyu Wu, Rongshan Gao, Maoliang Li, Zihao Zheng, Xinhao Sun, Hailong Zou, Guojie Luo, Xiang Chen
Abstract: Audio-driven video generation (A2V) has achieved promising progress in synthesizing temporally coherent and audio-visually aligned videos, yet its inference remains expensive due to the iterative denoising process of diffusion models. Existing caching methods mainly exploit temporal redundancy in visual features while overlooking the cross-modal alignment of A2V, where audio drives visual generation with highly non-uniform temporal importance. In this paper, we identify two levels of misalignment in existing A2V caching methods: temporal-semantic and computation-storage misalignment. To address them, we propose EchoCache, an energy-guided cross-modal caching framework for efficient A2V generation. EchoCache leverages audio time-frequency energy as a saliency anchor to guide latent-level cache updates and further introduces a dynamic timestep-latent caching mechanism with quantized cache management for joint efficiency and memory optimization. Extensive experiments on mainstream A2V models show that EchoCache consistently improves the latency-quality trade-off while preserving generation quality and audio-visual consistency. In particular, on Wan2.2-S2V over the EMTD benchmark, EchoCache achieves a 2.46x speedup with the best overall performance. Code is available at https://github.com/IF-LAB-PKU/EchoCache.
Authors: Vutichart Buranasiri, James M. Murphy
Abstract: Two active learning algorithms for hyperspectral image (HSI) classification are proposed that combine density-aware Fermat distances with Poisson-reweighted harmonic label propagation. Our methods actively query points using an uncertainty-based acquisition function, extending Poisson ReWeighted Laplace Learning (PWLL). Our first algorithm, Fermat Active Laplace Learning (FALL), builds an affinity matrix using Fermat distances between all data points. Then, PWLL is run with a diagonal perturbation using the minimum-norm acquisition function. In contrast, Approximate FALL (A-FALL) computes Fermat distances between each data point and landmark pixels selected via farthest-point sampling and constructs the affinity matrix using landmark multidimensional scaling. After several query rounds, A-FALL selects the Fermat exponent $p$ using a leave-one-out cross-validation variant. FALL and A-FALL leverage Fermat distances and subsequent harmonic label propagation to provide a density-aware estimation of the data manifold, improving labeling accuracy. Experiments on Salinas A and Pavia show the effectiveness of FALL and the scalability of A-FALL to large HSI scenes.
Authors: Zihan Yang, Yang Guo, Hongxing Zhang, Dan Lu, Siyuan Yao
Abstract: Despite the remarkable progress over the past decades, accurately identifying small objects remains challenging because of their insufficient visual cues. Previous works typically attempt to construct discriminative representation of the small objects. However, the wide range frequency domain noises and label ambiguities have been greatly overlooked, which significantly hinders the accurate localization. To address these issues, we propose a novel small object detection (SOD) detector termed DyFrDet, which is able to precisely localize the small object by dynamically suppressing the background distractions in frequency domain. Specifically, we propose a Dynamic Frequency-aware Feature Pyramid Network (DyFrFPN) to adaptively suppress low-frequency redundancy and excessive high-frequency noises. The DyFrFPN transforms the hierarchical features into frequency domain representation, and introduces a Dynamic Band Predictor (DBP) to preserve the discriminative components for small object identification. Afterwards, we present a novel Label Disambiguation Module (LDM), which leverages probabilistic distributions to explicitly model and alleviate the inherent ambiguity of target labels, yielding efficient improvement in localization precision of the small objects with low-resolution. Extensive experiments demonstrate that DyFrDet achieves state-of-the-art performance across multiple benchmarks, indicating its effectiveness and robustness in various challenging scenarios. Our code is available at https://github.com/ManOfStory/DyFrDet.
Authors: Jiayu Chen, Zhikun Jiang, Maoliang Li, Jiayi Luo, Jiawei Yang, Zihao Zheng, Hengyi Zhang, Guojie Luo, Xiang Chen
Abstract: Video Diffusion Transformers (VDiTs) enable high-fidelity generation but incur quadratic cost from dense 3D self-attention. Existing head- and block-level sparse methods share computation budgets across queries, overlooking token-specific attention demand. We observe that retained density varies across queries yet correlates log-linearly with attention entropy, while dominant interactions form query-centered neighborhoods with token-dependent radii. Based on these findings, we propose Token Radius Attention (TRA), a training-free framework that maps query entropy to an analytic token budget and converts it into a temporally decayed radius without explicit key ranking. Fused entropy extraction, warm-up reuse, and block-sparse mask construction further reduce overhead. Across seven Wan2.1, Wan2.2, and HunyuanVideo T2V/I2V configurations, TRA retains only 9-19% of attention interactions and achieves 1.56x-2.05x speedup with competitive generation quality. Code is available at https://github.com/IF-LAB-PKU/Token-Radius-Attention.
Authors: Nan Bi, Taoyue Wang, Lijun Yin, Vandana Sharma
Abstract: Automated pain assessment in real clinics is limited by scarce clinically grounded facial video data with weak labels (often sequence-level self-report) and by the fact that pain cues can be subtle or near-neutral in RGB, while thermal and depth signals are informative yet impractical to deploy routinely. To address these challenges, we propose ReMiX-MAE (Reconstructing Missing Channel Cross-Modal Masked Autoencoder), a self-supervised multimodal masked pretraining framework that learns transferable facial representations from synchronized RGB, thermal, and depth videos and explicitly trains robustness to missing modalities, enabling RGB-only deployment. To fill the gap of clinically grounded facial pain data with video-level self-report and longitudinal treatment trajectories, we collect the Sympathetic Mediated Pain (SMP) dataset with paired pre- and post-recordings across multiple visits. Under RGB-only deployment, we evaluate ReMiX-MAE using both direct feature extraction and pseudo-multimodal features decoded from RGB. ReMiX-MAE consistently outperforms an RGB-only masked autoencoder baseline on SMP, with pseudo-multimodal features providing additional gains in the challenging five-class setting. Across external datasets, ReMiX-MAE further shows more robust and label-efficient transfer than RGB-only baselines, highlighting its advantage in data-limited clinical settings.
Authors: Tingyu Song, Mingxin Li, Yanzhao Zhang, Dingkun Long, Pengjun Xie, Zhijie Nie, Yilun Zhao, Shu Wu
Abstract: Sparse retrieval underpins modern search systems, from web search to retrieval-augmented generation. Existing work has introduced Learned Sparse Retrieval (LSR) to push beyond exact lexical matching toward richer semantics. Yet LSR has so far remained tied to encoder-style bidirectional architectures, and its extension to multimodal settings still relies heavily on auxiliary cross-modal modules. To address these limitations, we introduce UEmbed (Unified Embedding), a decoder-only multimodal embedding model that produces both sparse lexical and dense representations in one causal forward pass. UEmbed appends N learnable special tokens to the input and partitions the vocabulary into N disjoint subsets. Each token's causal hidden state predicts sparse weights over its assigned subset, and the N subsets are concatenated into the full sparse vector. Trained on public data, we release UEmbed at 2B, 4B, and 9B scales. UEmbed-9B reaches 71.8 (dense) and 71.0 (sparse) on MMEB-v2, outperforming multimodal embedding models trained on publicly available data (e.g., RzenEmbed). On BEIR, UEmbed also remains competitive with strong dense and sparse baselines. Furthermore, we demonstrate the practical utility of UEmbed across three dimensions: effectiveness, efficiency, and agentic applications. Overall, UEmbed offers a new paradigm: it unifies dense and sparse embeddings in one model, while further extending sparse retrieval to unify text and multimodal inputs.
Authors: Zhipeng Liu, Haochen Wang, Zhaoxiang Zhang
Abstract: Captions serve as a primary supervision signal for both multimodal understanding and text-to-image generation. However, previous evaluations treat the caption quality as a single scalar objective, which conflates two distinct properties: (1) how much visual information a caption covers and (2) how reliably the image supports its stated claims. To this end, we design a decoupled caption evaluation benchmark, CAPEval (Coverage And Precision Evaluation), with human-written ground-truth captions and human-verified atomic checklist items. Specifically, CAPEval decomposes caption quality into Coverage and Precision. The former quantifies how thoroughly a caption covers ground-truth factual content, while the latter reflects the factual correctness rate of all claims expressed in the caption. We select 10 captioners and further conduct controlled downstream end-to-end experiments with them from four model families, where the caption source is the only variable. Empirically, we find a consistent task-dependent dissociation: Coverage serves as the stronger correlate for understanding performance, whereas Precision acts as the dominant predictor for generation performance. This decoupled evaluation paradigm not only delivers a more fine-grained diagnosis of caption quality, but also offers actionable guidance for selecting and optimizing captioners tailored to different downstream tasks.
Authors: Chao Ji, Shiyu Xuan, Zechao Li
Abstract: Aerial-ground person re-identification is a challenging task due to cross-platform viewpoint variations, which cause severe occlusion and geometric deformation. Existing methods attempt to learn view-invariant representations exclusively within the 2D image space, where drastic viewpoint variations cause the learned features to remain coupled with viewpoint bias. To address this, we propose VR3D, a View-Robust 3D Representation Learning framework that maps images into a unified 3D coordinate space to achieve view-independent feature interaction. Specifically, we introduce View-Robust 3D Representation Interaction, which leverages 3D priors extracted from single 2D observations to lift 2D appearance features into a canonical 3D space. VR3I employs 3D Geometry-Semantic Attention to establish interactions between 2D patches and 3D voxels from corresponding body parts based on their 3D spatial locations, effectively grounding 2D semantics within a 3D framework. In addition, as the reliability of these representations varies across samples due to viewpoint changes and 3D reconstruction errors, we introduce Reliability-Aware Fusion, which estimates sample-specific reliability and adaptively aggregates the multi-source representations. Extensive experiments on three benchmark datasets (CARGO, AG-ReID.v1, and AG-ReID.v2) demonstrate that VR3D outperforms recent methods. For example, it achieves a 5.63% improvement in Rank-1 on CARGO. Our code will be released.
Authors: Yuxue Yang, Shuyao Shang, Jiahe Wang, Zitong Zhou, Liang Tan, Junhan Zeng, Ruizhi Li, Junyan Li, Yu Liu, Xiao Yang, Yong Li, Jun Zhu, Hongsheng Li, Tieniu Tan, Lue Fan, Zhaoxiang Zhang
Abstract: Controllable video generation models are increasingly being developed as world models. Accordingly, evaluating them in this role extends beyond the apparent appearance of generated videos to the inherent reactivity of the worlds they depict: the ability to infer from the scene state how the world should react and to generate plausible consequences not explicitly described in the input. Yet existing benchmarks mainly assess visual quality or explicit instruction fulfillment by checking whether requested actions and interaction outcomes are realized, leaving inherent reactivity underexamined. We introduce WorldExam, a hierarchical diagnostic benchmark spanning four levels: Visual Quality, Control Adherence, Spatial Consistency, and World Reactivity. It comprises 1,474 cases across eight dedicated tasks and supports unified evaluation of camera-, action-, and language-driven model paradigms. The World Reactivity level evaluates scene-conditioned reactions and goal-directed behaviors beyond what is explicitly specified in the input. Evaluation of 20 representative models reveals a clear capability split. Camera-driven models excel at camera control, but their interfaces do not support dynamic interaction; action-driven models control subjects more precisely but often leave the world unresponsive; and language-driven models perform better on interaction but follow complex controls less faithfully. No model combines broad task coverage with consistently strong performance, showing that high visual quality and explicit instruction fulfillment do not guarantee inherent reactivity.
Authors: Ziran Li, Qiang Wang, Zhengyu Chen, Shanglin Lei, Borun Chen, Jingang Wang, Xunliang Cai
Abstract: Choosing the right large language model (LLM) backbone is the most consequential decision when building a vision-language model (VLM), yet it remains fundamentally unprincipled: compute-based scaling laws fail to generalize across model families, and no framework exists for directly predicting VLM performance before training begins. We propose the Capability-Driven Multimodal Scaling Law, the first cross-family framework that predicts VLM benchmark accuracy from directly observable textual capability. Given a low-dimensional capability score $S$ extracted from LLM textual benchmarks via PCA, we model VLM performance as a function of $S$, with a per-backbone transfer rate and an absorption rate that quantifies data-scaling efficiency. To fit and validate the framework, we train over 150 VLMs on 34 LLMs spanning 7 model families under a strictly controlled recipe. Evaluations on more than 200 textual and 50 multimodal benchmarks show that the law accurately extrapolates transfer rate from models up to 8B parameters to 72B-scale backbones, predicts full VLM training trajectories with high fidelity, and generalizes to entirely held-out model families. Beyond the scaling law, our analysis surfaces actionable insights: certain textual benchmarks negatively correlate with multimodal performance, exposing latent benchmark-gaming behavior; base LLMs outperform instruction-tuned counterparts as VLM backbones due to higher absorption rates and lower data-scaling decay; and different model families occupy distinct positions in the transfer--absorption space. The framework turns backbone selection from costly empirical sweeps into a principled, quantitative decision. Code and data are available at https://github.com/wangq-dev/CDMScaling.
Authors: Brian Song, Michael A. Lepori, Ellie Pavlick
Abstract: Goal-directed visual processing is a hallmark of human visual intelligence, resulting in representations that support downstream tasks such as categorization or search. Though vision-language models (VLMs) are often faced with these same tasks, their ability to recode visual representations when presented with goal-directed language remains poorly characterized. Indeed, prior work largely treats visual representations in VLMs as static repositories of visual information that are manipulated by language representations. In the present work, we provide evidence for two concrete instances of language-induced recoding of visual representations. First, we identify an abstract reference representation that denotes which objects are goal-relevant under a natural language prompt. We extract contrastive steering vectors corresponding to this reference representation and demonstrate that they are causally implicated in model predictions. These reference representations are abstract in that they generalize to different objects, different task contexts, and even from synthetic to naturalistic images. Second, we demonstrate language-induced attribute modulation: later layers selectively amplify goal-relevant attributes in visual representations of objects. We demonstrate this phenomenon across a range of different prompts. Finally, we provide a causal intervention that demonstrates that attribute modulation mediates a VLM's response distribution. Together, our results support a more dynamic account of cross-modality processing in VLMs -- rather than vision tokens serving as static repositories of information, they are modulated to support queries articulated in language.
Authors: Ali Al Housseini, Carlos Natalino, Paolo Monti, Omran Ayoub
Abstract: We propose a retrieval-based framework for crossdomain quality-of-transmission (QoT) estimation that leverages transferable feature representations while avoiding reliance on source-domain-specific decision boundaries. The proposed approach supports both zero-shot and few-shot adaptation without requiring model retraining. Experimental results on cross-domain QoT datasets demonstrate improved generalization performance compared with conventional machine learning baselines and recent contrastive learning approaches, highlighting the potential of retrieval-based inference for robust optical network automation.
Authors: Shiwen An, Zhongyi Ni, Huanhai Zhou, Jin-Guo Liu
Abstract: The Discrete Fourier Transform, the Discrete Cosine Transform, and their block-wise variants underpin most deployed image and video codecs. Their effectiveness rests on three properties: they run in near-linear time (linear up to a polylogarithmic factor), they are exactly invertible, and they carry few to no parameters. In this work, we generalize these bases to isometric multilinear bases, allowing a small number of extra parameters, polylogarithmic in the image size, while preserving all three properties. Given an image dataset, we develop a systematic framework that searches this family for the basis compressing the dataset most effectively: the basis is parameterized as an isometric tensor network, inspired by quantum many-body theory, and trained with Riemannian optimization on the manifold of unitary matrices. Across natural photographs and line drawings, the trained bases consistently improve on their fixed, non-parametric counterparts. On Quick Draw line-drawing compression, they store images in roughly $20\%$ fewer bytes than JPEG's $8 \times 8$ block cosine transform at the same reconstruction quality.
Authors: Kuan-Yen Chen, Fang-Yi Su, Philip Chikontwe, Jung-Hsien Chiang
Abstract: Image restoration quality can be evaluated along three complementary facets: pixel-level fidelity, human perception, and downstream machine preference. However, existing lossy compression restoration methods optimize for at most one of these criteria: fidelity-oriented models often regress toward conditional means and produce over-smoothed outputs, while generative approaches hallucinate plausible but factually incorrect textures that degrade both ground-truth fidelity and downstream task accuracy. To navigate this three-way tradeoff, we propose FDIR, a two-stage architecture that decouples the conflicting demands through complementary inductive biases: Quality-Guided One-Step Flow Matching (QO-Flow) recovers global semantic structure in latent space via a single forward pass, while Flow-Conditioned Detail Refinement (FCDR) deterministically restores high-frequency textures and suppresses generative hallucinations in pixel space. Extensive experiments demonstrate that FDIR achieves superior fidelity, with a favorable perceptual-fidelity balance and competitive machine preference.
Authors: Alexandros Haridis, Charles Zhou
Abstract: Design and architectural archives encode expert human knowledge in graphical formats, providing a critical testbed for design-inspired Machine Learning (ML) challenges absent with typical computer vision benchmarks. Building on JONES-19, a small-size image dataset based on The Grammar of Ornament (London, 1857), we evaluate the discriminative performance of Convolutional Neural Networks (CNNs) in two model training strategies: (a) ImageNet pretraining for domain-general "visual common sense," and (b) learning from scratch on the design data in JONES-19. We find that while domain-general priors improve discriminative performance, learning from scratch augmented with repeated local sampling (multi-crop) effectively recovers these gains. For highly structured design data, local design-driven representations provide sufficient foundation for learning, challenging a reliance on massive general-purpose pretraining. These findings suggest that in specialized design domains, careful curation of smaller high-quality datasets that capture empirical and formal design principles may prove more effective and informative on the nature of a particular design domain than prioritizing large-scale data collection.
Authors: Yifan Chen, Jiaming Liu, Ye Zheng, Zhe Sun, Tao Chen
Abstract: Underwater image enhancement (UIE) aims to recover clear images from observations affected by wavelength-dependent absorption, scattering, and spatially nonuniform degradation. Although existing generative methods can handle complex degradations, severe information loss may lead to semantic drift in the restored results. To address this issue, we propose RPL-UIE, a two-stage teacher--student framework for reliable prior learning. In the teacher stage, the network learns reliable and complementary spatial priors characterizing appearance and photometric properties from paired degraded and reference images. In the student stage, the network takes only degraded images as input and learns to emulate the teacher's prior extraction capability, thereby providing more reliable restoration guidance for the enhancement process without requiring reference images at inference. To reduce the prior-learning discrepancy between the teacher and student models, we further develop Residual Prior Refinement Diffusion (RPRD) and Frequency-Aware Prior Residual Calibration (FPRC). RPRD uses the coarse priors as anchors and progressively predicts the necessary corrections in the residual space. FPRC retains stable low-frequency residual components and selectively modulates high-frequency detail residuals, producing calibrated priors to support high-quality reconstruction. Experiments on multiple UIE benchmarks demonstrate competitive restoration performance. Downstream underwater object detection and instance segmentation experiments further demonstrate the improved utility of enhanced images for visual perception, while tests on real-world data captured by a remotely operated vehicle (ROV) support the practical applicability of RPL-UIE.
Authors: Yi Yu, Yixuan Liu, Ziyu Zhang, Parker Martin, Zhenyu Bu, Yuchi Han, Yuan Xue
Abstract: Accurate estimation of left ventricular (LV) orientation is essential for cardiac magnetic resonance (CMR) imaging and downstream analysis. Existing methods typically formulate orientation recognition as discrete view classification or rely on multi-slice geometric intersection, limiting their ability to model continuous 3D orientation and generalize across arbitrary slices. This work introduces a novel paradigm: Joint LV localization and 3D orientation estimation from a single CMR slice. To investigate this setting, representative orientation-aware detection frameworks are adapted to the CMR domain, and their limitations are analyzed. Upon that, we propose the Polar-Coupled Circular (PCC) embedding that provides a continuous and unambiguous orientation representation to address the limitations. Meanwhile, a scalable benchmark is constructed through automatic slice sampling from volumetric CMR segmentation datasets. Extensive experiments on four datasets demonstrate strong performance, achieving an average mIoU of 86.18% and an average angle deviation of 3.39{\deg}. This study establishes a new task setting for single-slice LV orientation modeling and provides a geometry-consistent framework for spatially informed CMR analysis. Code is available at https://github.com/yuyi1005/cmr-3d-ood.
Authors: John Garcia Henao, Nicholas B\"unger, Benedikt Herzog, Cindy Guerrero Toro, Benjamin Vella, Matthias Biner, Rico Br\"utsch, Carmen Castroviejo Fernandez, Felix \"Ottl, Norman Juchler, Armando Hoch, Bettina Hochreiter, Sven Hirsch, Sebastiano Caprara
Abstract: High-resolution 3D segmentation of hip and shoulder anatomy from CT and MRI is essential for surgical planning, yet frozen segmentation models often fail under domain shift. CNN-based expert models are fully automatic but lack adaptability, whereas promptable foundation models generalize better but require manual prompting. We present MedSAM2-Anatomy, a training-free inference-time optimization framework that improves frozen segmentation models without retraining or human interaction. A frozen expert model generates anatomical priors that are automatically converted into multiple prompt hypotheses for a frozen 3D foundation model. Candidate masks are fused while anatomically implausible priors are rejected. No model weights are updated and no manual prompts are required. TotalSegmentator and MedSAM2 are used as representative expert and foundation models, allowing the contribution of the inference policy to be isolated. Evaluation on the independent Balgrist-V0 CT and MRI cohorts shows that inference-time optimization increases median Dice from 0.71 to 0.92 on hip MRI and from 0.89 to 0.92 on shoulder CT, while reducing median HD95 on hip MRI from 22.0 mm to 5.0 mm. On public TotalSegmentator benchmarks, the expert model remains strongest, indicating that the optimal fusion strategy depends on the reliability of the expert prior. These results demonstrate that training-free inference-time optimization provides a practical strategy for improving frozen segmentation models without manual prompting.
Authors: Kejia Zhang, Youran Sun, Chugang Yi, Haizhao Yang
Abstract: Do frontier vision-language models present a coherent Theory-of-Mind (ToM) profile across tasks, matching the same human reference group, or does that profile fragment from one paradigm to the next? We evaluate a shared panel of nine frontier VLMs on two psychology-derived benchmarks: the Keysar Director Task (visual perspective-taking under egocentric interference) and the Frith-Happ\'e animated triangles scored with the Castelli rubric (intention attribution from pure motion). On the Director Task, without chain-of-thought, the panel makes the egocentric error on 78\% of trials like children rather than adults; variation is substantial across models, and reasoning rescues several models. On the triangles, the panel under-attributes intention: its ToM profile sits more than three times closer to the high-functioning-autistic-adult (HF-ASD) mean than to the typical-development-adult (TD) mean, while Goal-Directed and Random stay near TD. No model is nearest TD on both tasks; the model that looks adult-like on the Director Task falls on the HF-ASD side on the triangles, and the most TD-like model on the triangles is child-like on the Director Task. We report group-level descriptions, not diagnostic labels for any model.
Authors: Xuan Cuong Ngo
Abstract: Recent radiology-adapted vision-language models have achieved strong performance on standard report generation benchmarks, yet their robustness and generalization remain constrained by imperfect alignment and correlation between visual and textual features. Existing methods connect image and text either implicitly through autoregressive report supervision or explicitly through contrastive learning. However, autoregressive supervision alone is insufficient to establish reliable image-text alignment, while contrastive learning can push apart unpaired reports that describe related pathologies simply because they are not paired with the same image. This is problematic in radiology, where different reports may share compatible pathology semantics rather than being true negatives. As a result, the learned representation may fail to organize images and reports around shared pathology concepts, causing the decoder to rely on pretrained language priors and generate clinically plausible reports that are not fully supported by radiographic evidence. To address this issue, we propose PALM, a pathology-aware alignment framework for radiology report generation. Instead of directly matching each image-report pair while separating all others, PALM aligns visual and textual features through shared pathology prototypes. These prototypes provide a clinically meaningful bridge between radiographic evidence and textual findings, allowing cases with similar pathology semantics to move toward common concepts without separating compatible cases. In addition, we introduce Masked Evidence Modeling to strengthen the image encoder sensitivity to local radiographic evidence by learning semantic changes caused by masked image regions. Experiments on MIMIC-CXR, IU X-Ray, and MIMIC-ABN show that PALM consistently improves both report generation and abnormality-focused robustness.
Authors: Sparsh Roy, Vihan Aggarwal, Davin Yin
Abstract: A robot that has to see and react on a fixed clock runs into two problems at once. Its cameras degrade in rain, mud, fog, and darkness. And the single onboard processor it runs on is shared with planning and control, so the compute left over for perception moves around from second to second. Most systems model the two separately. We present a perception router that tracks probabilistic estimates of sensor-fault state and compute- contention state, couples them with a noisy-OR term, and uses the coupled estimate to pick one of four detector configurations (YOLO11x/n at 1280 or 640 px) so that the frame finishes before its deadline. Where the two stressors co-occur, the coupled policy cuts the deadline-miss rate by 1.1 to 9.4 percentage points against a policy that treats them independently. The interval excludes zero in five of six conditions, the pooled effect over 10 sequences and 6 conditions has sign-test p = 0.001, and every uncoupled control and the fault-free trajectory sit at exactly 0.0 pp. Routing costs tens of microseconds per frame. We then asked whether the coupling the method exploits arises on its own. Across eight real RADIATE adverse-weather sequences and three workload proxies independent of the fault signal, after Benjamini-Hochberg correction and a replication run, none of 24 tests found it. We report that null and scope the routing result as a proof of mechanism. Whether such coupling occurs in the field is still open, and the released evaluation pipeline lets a deployment settle it on its own traces.
Authors: Jasin Cekinmez, Addison J. Wu, Raja Marjieh, Thomas L. Griffiths
Abstract: Human language is highly polysemous. Many common words (e.g., 'bank' or 'palm') carry several distinct meanings that shape what humans communicate and imagine. Large language models (LLMs) have been shown to understand this multiplicity of meaning, but much less is known about how polysemy surfaces in other modalities such as images. We study this across 17 text-to-image and 15 text-generation models by giving each a polysemous word with no context to fix its meaning and measuring which senses are produced over many samples. We find a clear multimodal gap, where within every model family, generated images settle on far fewer senses than generated sentences (normalized entropy 0.10 vs. 0.25), and both are far less varied than what people imagine for the same words (normalized entropy 0.47). However, when we instead ask a model to list how often it would generate outputs corresponding to each possible meaning of a word, it predicts distributions that are more diverse than the actual space of outputs. These results reveal a multimodal gap in how foundation models express meaning, and how their understanding may not transfer faithfully nor equally across modalities.
Authors: Jinyuan Liu, Tianshuo Cong, Pei Li, Tianrui Wang, Xinlei He, Anyu Wang, Xiaoyun Wang
Abstract: Although semantic watermarking is considered a promising safeguard for images generated by Latent Diffusion Models (LDMs), the reliance of the watermark detection pipeline on neural networks introduces a critical yet underexplored backdoor attack surface. To systematically study this vulnerability, we propose GhostVAE to plant a stealthy backdoor into the encoder of Variational Autoencoder (VAE), enabling reliable evasion of watermark detection. GhostVAE operates in two stages: it first constructs a universal trigger via power spectrum regularization to improve the trigger robustness, and then trains a backdoored VAE encoder with a parameter-aligned objective. Through extensive evaluations across three state-of-the-art semantic watermarking schemes and three widely adopted LDMs, we show that GhostVAE preserves watermark detection performance on benign images (achieving an average true positive rate of 94.4%), while simultaneously enabling highly effective evasion under trigger activation (achieving an average attack success rate of 94.6%). Moreover, we comprehensively analyze seventeen representative defenses and demonstrate that GhostVAE remains stealthy across the input space, parameter space, and latent space. Our work fundamentally undermines the trustworthiness of semantic watermarking systems and highlights that secure deployment of semantic watermarks requires end-to-end security considerations, particularly for neural network components.
Authors: Xiaoyang Chen, Shengcheng Luo, Haoran Guo, Jiaming Jiang, Wanlin Li, Ziyuan Jiao, Chenxi Xiao
Abstract: Dexterous object rotation is a sequential contact problem: each support, release, and re-contact decision must both produce the desired object motion, and prepare the hand configuration for continued rotation. Existing reinforcement learning methods discover such movement patterns through trial and error on specific robotic hand embodiments, without explicitly accounting for how each contact transition affects the hand's ability to sustain object rotation in subsequent steps. We introduce DexMani, a framework that transfers human demonstrations as contact-conditioned manipulability evolution. This prior captures how successful human contact transitions reshape the object-rotation directions available to the hand. DexMani then learns this manipulability evolution and uses it to guide downstream reinforcement learning, enabling rotation skills to be acquired across robot embodiments with distinct kinematics and active-contact configurations. Across the Shadow Hand, Allegro Hand, and XHand, DexMani achieves the highest success rates in every evaluated setting for both seen and unseen objects. DexMani reaches an average success rate of 57.5% on LEAP Hand, outperforming other baselines and producing smoother rotatory motions. Project site: https://dexmani.github.io
Authors: Shalom Kachko, Raz Lapid, Margarita Vald, Almog Dubin, Moshe Sipper
Abstract: Vision-language models (VLMs) process image patches and text tokens in a shared residual stream, but the local geometry through which the two modalities interact remains poorly understood. Most interpretability methods identify global linear directions, which may miss representations that are globally high-dimensional but locally low-dimensional. We introduce LENS (Local Explanation of Neighborhood Subspaces), a method that decomposes VLM activations into local low-rank Gaussian neighborhoods using a Mixture of Factor Analyzers. Applied to LLaVA-1.5-7B and Qwen3-VL-8B, LENS reveals distinct depth-dependent fusion trajectories consistent with each model's fusion mechanism: LLaVA progressively mixes modalities at later layers, whereas Qwen3-VL mixes them early, partially re-segregates them, and recombines them near the output. An automated multimodal labeling pipeline assigns concise semantic descriptions to these neighborhoods. Interpolating activations toward neighborhood centroids causally redirects generation within and across modalities and outperforms difference-in-means and VL-SAE in most evaluated conditions; in one LLaVA vision-to-vision setting, MFA achieves 5.7 times the VL-SAE score. Human evaluation finds MFA steering competitive with prompting and substantially stronger than the other intervention baselines. Finally, the MFA coefficient space improves Qwen3-VL image-to-rendered-text retrieval at the deepest evaluated layer from 14.9% to 48.6% R@1. Ablations show that the reported fusion trajectories are stable across component counts, local ranks, and modality-purity thresholds. These results support local geometric neighborhoods as useful interpretable and causal units for analyzing cross-modal representations in the evaluated VLMs.
Authors: Zixuan Zhu, Rui Wang, Lihua Jing, Jinwen Zhong
Abstract: Backdoor attacks pose a serious threat to deep neural networks, especially when training relies on third-party data, allowing adversaries to inject malicious behaviors through data poisoning. In this work, we reveal that backdoor behaviors tend to be absorbed by a simpler parallel branch when jointly trained with the main network. Motivated by this insight, we propose Trapping and Removing (TR), a simple yet effective training-time defense that introduces a lightweight shortcut branch as a "honeypot" to trap backdoor knowledge. After training, backdoors can be removed by discarding the shortcut, without requiring any additional data. To further enhance backdoor isolation while maintaining benign performance, we design a knowledge decoupling strategy with entropy-based weight assignment, encouraging poisoned samples to flow through the honeypot while guiding the main network to focus on benign learning. In addition, we introduce an automatic shortcut generation strategy to improve generalization across model architectures. Extensive experiments on four benchmark datasets and five model architectures demonstrate that our approach effectively mitigates a wide range of backdoor attacks while preserving performance on benign data. Code: https://github.com/Zixuan-Zhu/TR}{github.com/Zixuan-Zhu/TR.
Authors: Ruoyu Wu, Zhenhong Sun, Xiaoming Gong, Yuxin Xian, Zhi Wang, Yawen Chen, Huadong Mo, Daoyi Dong
Abstract: Accurate 3D part decomposition requires separating shapes into structurally meaningful components with precise boundaries while preserving articulation seams and thin attachments. Existing approaches often suffer from a structural-scale mismatch: geometric evidence for separation is most reliable at the meso scale, yet many pipelines operate either too globally to respect joints or too locally to remain robust to noise. We propose Hi-TOPS, a Hierarchical Topology-aware Scoring Prior that aggregates complementary intrinsic cues into a multi-resolution Flow-Freeze field. Flow regions provide expandable support for primitive coverage, while Freeze regions restrict growth near articulations and thin structures. A TSDF-guided body-surface superquadric fitter then captures dominant cores and residual surface structures, followed by SQ-to-mesh assignment for connected, boundary-aligned parts. Across diverse benchmarks, Hi-TOPS delivers stable, editable decompositions without semantic supervision or 2D foundation priors.
Authors: Ivan Snegirev, Elizaveta Semenyakina, Mikhail Konenkov, Artem Lykov, Miguel Altamirano Cabrera, Dzmitry Tsetserukou
Abstract: ORCESTRA is a mixed-reality system for programming robot digital twins through no-code waypoint teaching and language-guided control. In a passthrough mixed-reality workspace, users place robot twins on real surfaces, teach trajectories, save robot-relative episodes, or issue spoken/typed commands that a vision-language model converts into structured digital-twin plans. Both interaction modes share a backend for metric grounding, embodiment-aware validation, preview, confirmation, and digital-twin execution. The system supports heterogeneous robot embodiments, including fixed-base manipulators, a mobile base, and a humanoid robot, demonstrating MR validation as a safety layer for language-guided robot programming before physical deployment.
Authors: Zag ElSayed, Payton Siekierski, Jack Yanchen Liu, Ernest Pedapati
Abstract: Fragile X Syndrome (FXS) is a neurodevelopmental disorder caused by reduced expression of fragile X mental retardation protein (FMRP), leading to disrupted synaptic plasticity, cortical hyperexcitability, and impaired network synchronization. Electroencephalography (EEG) provides a noninvasive window into these mechanisms and consistently reveals abnormalities in alpha (8 to 12 Hz) and gamma (30 to 100 Hz) oscillations that relate to inhibitory control, sensory processing, and cognition. This paper proposes a multi representation deep learning framework for automated characterization of FXS EEG phenotypes by integrating convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and recurrence plot (RP) analysis. Band limited EEG signals are decomposed into alpha and gamma components and transformed into complementary representations, including temporal feature sequences, time frequency maps, and RP images encoding the nonlinear recurrence structure. CNN modules learn discriminative spatial-spectral and dynamical textures from image based representations, while LSTM modules model temporal modulation of oscillatory activity; a hybrid CNN LSTM architecture jointly captures spatial, temporal, and nonlinear dependencies. Subject-independent evaluation demonstrates that the hybrid model outperforms single modality baselines, with gamma features providing strong discriminative power and alpha gamma integration yielding the best overall performance. These findings support deep learning with nonlinear representations as a scalable approach for EEG biomarker development in FXS, with potential utility for diagnosis, stratification, and treatment monitoring in translational settings.
Authors: Muhammad Irfan Khan, Eero Lehtonen, Joni Obradovic, Elina Kontio, Esa Alhoniemi, Suleiman A. Khan, Mojtaba Jafaritadi
Abstract: Federated Learning (FL) enables collaborative training of machine learning models across multiple institutions without sharing sensitive data, making it particularly suitable for medical imaging applications. However, heterogeneous data distributions across institutions and potential information leakage through model updates remain important challenges. In this work, we propose DP-SimAgg, a privacy-preserving federated learning framework that integrates similarity-weighted aggregation with a server-side differential privacy mechanism. The proposed method applies L2 clipping to bound collaborator updates, computes similarity-based aggregation weights to mitigate the effects of non-IID data distributions, and injects calibrated Gaussian noise at the central server, providing per-round privacy guarantees under the assumed sensitivity bound. The framework is implemented using Intel's OpenFL platform and evaluated on the FeTS 2022 dataset consisting of 1251 multi-modal MRI scans for brain tumor segmentation. Experimental results demonstrate that DP-SimAgg maintains competitive segmentation performance while providing privacy protection. Under a strict per-round privacy budget (epsilon = 1, cumulative epsilon_total = 20 over 20 rounds), the method achieves Dice scores of 0.6357, 0.5305, and 0.5274 for the enhancing tumor (ET), tumor core (TC), and whole tumor (WT) regions, respectively. With a more relaxed per-round budget (epsilon = 10, cumulative epsilon_total = 200), performance approaches that of the non-private baseline while incorporating a central Gaussian mechanism with per-round (epsilon, delta)-DP accounting under the assumed sensitivity bound. These results highlight the potential of DP-SimAgg for enabling privacy-preserving collaborative learning in medical imaging applications.
Authors: Kilian Northoff, Mateo de Mayo, Daniel Cremers
Abstract: 3D Gaussian Splatting (3DGS) provides efficient rendering of photo-realistic scenes, but its heavy preprocessing and training steps make it a poor fit for applications that require real-time reconstruction in robotics or XR. This capability is important since it allows immediate feedback and interaction with new environments. Visual-inertial odometry (VIO) and simultaneous localization and mapping (VI-SLAM) systems, on the other hand, specifically target these real-time applications, which makes them a good choice for integration with 3DGS. We propose a new method that tracks and reconstructs simultaneously in real-time by leveraging an efficient visual-inertial tracking system based on Basalt together with a novel incremental method built on top of Brush, an efficient Rust-based GPU-vendor-agnostic implementation of 3D Gaussian Splatting. We show that many of the heavy preprocessing and training steps of 3DGS can be replaced with a more efficient incremental training strategy that has direct access to the information generated by the visual-inertial tracking system. Furthermore, we propose and combine multiple practical improvements to increase the efficiency of the training pipeline and adapt it to run in real-time, parallel to the tracking thread. This work highlights the value of exploiting the complementary nature of SLAM and 3DGS, and how that can lead to promising results for real-time 3D reconstruction.
Authors: Jibao Yuan, Yuhui Zhao, Yinzhen Lv, Chao Xu, Shun Li, Chenxi Deng, Shaofei Chen
Abstract: Existing 6-DoF grasp detectors typically rank grasp candidates by detector confidence. However, our analysis on GraspNet-1Billion shows that detector confidence is often poorly aligned with grasp quality, causing successful grasp candidates to be ranked too low during execution. Motivated by this observation, we formulate grasp candidate re-ranking as a separate task for frozen detectors, aiming to improve candidate ordering without changing the detector or its grasp candidates. We propose GraRe, which estimates grasp quality from candidate attributes, shell-stratified local geometry, and object context. Candidate attributes condition the local geometric and object-context representations, and a Transformer fuses all three feature types. The predicted quality is combined with detector confidence to produce the final ranking. Experiments on GraspNet-1Billion with three frozen detectors show consistent improvements, with gains of up to 13.60 points in Average AP. Real-robot experiments further demonstrate robust grasping in cluttered scenes. These results show that improving candidate ranking provides a practical way to enhance frozen 6-DoF grasp detectors.
Authors: Jin-Young Kim, So-Yoon Cho, Hyun-Gyoon Kim
Abstract: Flow Matching trains continuous-time generative models by regressing the velocity field of a probability path between a simple source distribution and a target data distribution. The coupling that pairs source and target samples strongly affects optimization and sample quality, but structured couplings typically rely on mini-batch transport or assignment procedures whose cost grows at least quadratically in batch size. We propose Quantile Coupling Flow Matching (QC-FM), a lightweight one-sided coupling: rather than matching two pre-sampled batches, it samples only the data batch and constructs each paired source directly. Data ranks projected along a small number of random orthogonal directions are mapped to Gaussian quantiles, and the latent code is completed in the orthogonal complement by conditional Gaussian sampling. The construction is one-dimensional per slice, so the coupling requires no pairwise cost matrix and no assignment to solve. We show that, for each drawn frame, this coupling eliminates the irreducible regression variance along every selected slice and makes the ideal flow exactly straight there, while leaving the sampling prior unchanged: generation still starts from the standard Gaussian, and the training source deviates from it only through the copula of the slice codes, whose transport cost we bound. For training, we apply QC to an anchor subset and complete the remaining source slots with exact Gaussian samples, retaining the QC bias while preserving an explicit signal from the Baseline coupling. Across CIFAR-10, CelebA, FFHQ, and ImageNet-64, QC-FM improves over the Baseline under matched training budgets, reducing FID by up to 12.9%, and outperforms OT-CFM on all four datasets. These results suggest that preserving projected rank structure is a simple and scalable way to inject useful geometric bias into FM couplings without solving a mini-batch transport problem.
Authors: Zhihao Zhu, Hanlin Shang, Mingwang Xu, Feipeng Cai, Zhuolin He, Yaoyi Li, Jianhua Han, Hang Xu, Siyu Zhu
Abstract: Vision-Language-Action (VLA) models have emerged as a prominent paradigm for end-to-end autonomous driving; however, their efficient deployment is severely constrained by high computational latency and exposure bias arising from sequential autoregressive decoding. Conversely, while specialized diffusion policies enable low-latency, parallel execution, training them from scratch typically yields narrow, single-task architectures that lack holistic visual-linguistic reasoning. Successfully transforming pre-trained autoregressive generalists into parallel diffusion models could combine multi-task cognitive intelligence with execution efficiency, yet this transition presents a formidable architectural challenge due to mismatched attention patterns (causal versus bidirectional) and divergent optimization objectives. To bridge this divide, we introduce WAM-Diff2, a multi-task discrete diffusion VLA framework powered by a three-stage hierarchical distillation strategy. By structuring the architectural shift through progressive block-wise adaptation, block-wise distillation, and model-wise cross-scale distillation, WAM-Diff2 preserves the underlying semantic foundations of the base model while accelerating inference. Extensive evaluations across driving understanding, perception, and planning benchmarks demonstrate that WAM-Diff2 effectively mitigates exposure bias and achieves performance parity with autoregressive baselines. Crucially, the autoregressive-to-diffusion transition yields a 2.8x decoding speedup, which scales to an ultimate 15.1x acceleration when combined with system-level optimizations including FlashInfer and CUDA Graphs.
Authors: Kapil Wanaskar, Gaytri Jena, Aman Chadha, Vinija Jain, Vasu Sharma, Amitava Das
Abstract: World models have attracted significant attention for their ability to capture and predict the structure and dynamics of the physical world. In this emerging landscape, Joint Embedding Predictive Architectures (JEPA) offer a particularly compelling direction. We study a largely unexplored regime: populous, crowded, and chaotic Global South urban environments, which we call DENSEWORLD. Unlike the lower-density, lane-structured settings that dominate existing evaluations, these scenes exhibit soft spatial boundaries, extreme agent heterogeneity, persistent occlusion, and rapid social negotiation under mixed traffic. We introduce the first large-scale dataset for this regime: 1,000 hours of drive-through, walk-through, and aerial video across 22 cities. Existing JEPA formulations struggle to preserve dense interaction dynamics under heterogeneity and partial observability. We introduce FactorJEPA, which makes world structure a first-class predictive primitive. Rather than encoding the future in a monolithic latent, it composes layout, entities, and interactions, using a visibility gate and separated subspaces to preserve partially observed agents and discourage cross-factor shortcuts. FactorJEPA improves (i) future-latent accuracy (Future-frame L1), (ii) intervention-sensitive prediction (Causal L1), and (iii) robustness to reduced visual evidence (Mask-ratio slope), while exposing (iv) a reproducible motion-information trade-off (Motion cosine). Method rankings replicate across 2B and 1B V-JEPA 2.1 backbones, with rho = 0.895 to 0.978. We publicly release the DENSEWORLD-115k dataset (https://huggingface.co/datasets/anonymousML123/denseworld-115k) and the surgery-trained FactorJEPA checkpoints (https://huggingface.co/datasets/anonymousML123/factorjepa-outputs/tree/main/outputs/full/vjepa_2_1_vitg_1B/train/m09c_surgery_3stage_DI_diheavy_encoder).
URLs: https://huggingface.co/datasets/anonymousML123/denseworld-115k), https://huggingface.co/datasets/anonymousML123/factorjepa-outputs/tree/main/outputs/full/vjepa_2_1_vitg_1B/train/m09c_surgery_3stage_DI_diheavy_encoder).
Authors: Mayank Sharma, Rohit Kumar Mourya, Pratik Mazumder
Abstract: Tabular data is used extensively in many real-world use cases. Deep learning models have been developed to deal with tabular data, but generally perform poorly when the test data distribution differs from that of the training data. Researchers have proposed test-time adaptation approaches to deal with this problem. The fully test-time adaptation (FTTA) setting involves adapting deployed classifiers to shifted target distributions using only unlabeled test data. Leading FTTA methods inherit a batch-dependent approach from computer vision literature. This paper demonstrates for the first time that such approaches degrade sharply in strict streaming regimes where examples arrive and must be classified one at a time. This occurs because at a batch size of one, batch-level statistics become unavailable or poorly estimated. We argue that singleton tabular FTTA is not merely a small-batch variant of ordinary FTTA, but a distinct identifiability problem where only the location of the model's score stream remains directly observable. To address this, we propose Prequential Logit-Origin Centering (PLOC), a lightweight approach that keeps the source model frozen and shifts the logit space at each step. PLOC stores only a single running number (the mean of past logits), requires no labels, estimates no priors, and bypasses weight updates entirely. A deferred variant applies a static shift that preserves the source ranking, and thus the AUROC, exactly. Evaluated across five tabular benchmarks, three architectures (MLP, FT-Transformer, and TabTransformer), and five independent source checkpoints, PLOC significantly outperforms strong tabular and entropy-based baselines.
Authors: Ganghyeon Lee, Inha Lee, Junhee Lee, Jeongeon Lee, Sung Whan Yoon, Kyungdon Joo
Abstract: Although recent robot perception research emphasizes training on data from diverse environments to improve generalization, most existing methods still rely on centralized learning, which is inefficient and difficult to scale across heterogeneous robot platforms. Federated learning (FL) offers an alternative by enabling distributed training without raw data transfer, but it suffers from severe performance degradation under domain shifts caused by heterogeneity across clients. In real robotic deployments, data distributions often overlap across platforms, environments, and sensing conditions, making it difficult to partition clients into clearly separated domains. However, this characteristic breaks the assumption of clearly separable client domains commonly used in clustered FL. To address this gap in robot perception, particularly in depth estimation, we introduce two realistic and unexplored non-IID scenarios that reflect heterogeneity in terms of platform, environment, and depth distribution. We then propose FeDepth, a descriptor-based clustered FL framework that models client relationships through soft clustering. Unlike hard clustering methods that assume clearly separated clusters, FeDepth allows clients to participate in multiple clusters, capturing continuous and ambiguous domain transitions commonly observed in robotic environments. Extensive experiments demonstrate that FeDepth consistently improves robustness over standard FL and clustered FL baselines across multiple depth estimation architectures, providing a practical and effective solution for federated robot perception. Our project page is available at https://vision3d-lab.github.io/fedepth/.
Authors: Ganzhong Luo, Yang Ren, Hanyong Wang, Shuyu Zheng, Menglong Yang
Abstract: Knowledge-Based Visual Question Answering (KB-VQA) requires retrieving relevant entity knowledge from external sources to answer visually grounded questions. Existing retrieval-augmented systems suffer from two critical limitations. First, relying on a single retrieval modality creates a Single-Source Retrieval Bottleneck, missing ground-truth entities that are only accessible through complementary sources. Second, dual-tower pointwise rerankers suffer from Retrieval-Source-Blind Reranking, as they overlook retrieval origins and candidate-level retrieval priors, leading to redundant modality reliance. To address these challenges, we propose UniHEAR, a unified lightweight framework for heterogeneous-source entity retrieval and reranking. UniHEAR constructs a Coarse Retrieval Descriptor for each candidate entity, and introduces Retrieval-Guided Attentive Modality Gating to condition modality attention weights on this descriptor, further complemented by Entropy-Weighted Source Fusion of coarse retrieval priors. A hybrid training strategy combining contrastive learning with an auxiliary modality-preserving loss unifies entity-level and section-level retrieval within a single model. Extensive experiments on E-VQA and InfoSeek demonstrate that UniHEAR achieves state-of-the-art retrieval and VQA performance, improving Recall@1 by 6.7 and 1.2 points over the strongest baselines while maintaining a lightweight reranking architecture. Code and model are available at https://github.com/iven-luo/UniHEAR.
Authors: Volodymyr Havrylov, Faris Janjo\v{s}, Andreas Look, J\"urgen Mathes, Andreas Geiger
Abstract: End-to-end autonomous driving (E2E AD) systems integrate perception, prediction, and planning into a single differentiable architecture. While these models show great promise, their standard training often relies on output-only supervision, which can lead to weak gradients for the hidden layers of increasingly complex models. Recent works have integrated vision-language model (VLM) supervision for latent features to address this, yielding substantial empirical gains, yet leaving the underlying theoretical mechanisms poorly understood. Our investigation into this methodology reveals that the resulting performance gains stem not from VLM reasoning capabilities, as previously assumed, but rather from the latent connections forged between the E2E AD model and ground-truth (GT) data during training. Building on this insight, we propose a probabilistic deep supervision framework that regularizes intermediate latent representations directly from GT data. By treating model latents as reparameterizable distributions, we optimize the architecture via the Evidence Lower Bound (ELBO). Our evaluations conducted on the nuScenes dataset demonstrate that supervising trajectory-related latents with future GT paths consistently improves planning performance. Using identical training data and E2E architectures, our method achieves an 8% reduction in planning L2 error and a 3% decrease in collision rates compared to competitive vectorized baselines, all while incurring negligible computational overhead.
Authors: Qi Lv, Jianming Xing, Zhao Yang, Mingyuan Yao, Yinan Shi, Yawei Jueluo, Mike Zheng Shou, Xiang Deng
Abstract: Despite recent progress in Vision-Language-Action (VLA) models for robotic manipulation, the action chunk remains a weakly structured interface. Existing work typically flatten each chunk into per-timestep controls, relying on implicit data learning that manifests as jagged motion and boundary discontinuities during physical execution. To address these limitations, we introduce Hermite trajectory priors, parameterizing the chunk trajectory as a piecewise cubic Hermite curve defined by endpoint positions and velocities to explicitly enforce smoothness and continuity. We instantiate this fixed operator across discrete autoregressive and continuous generative paradigms via three variants: (1) Hermite Tokens, which predict quantized boundary variables autoregressively; (2) Hermite Scaffold, which decomposes clean actions into a base scaffold and residuals; and (3) Hermite Regularization, which applies the prior strictly as an auxiliary training objective. Across simulation benchmarks and real-robot platforms, Hermite Regularization achieves superior performance among these three variants, improving {\pi}0.5 baseline success rates from 95.9% to 98.7% on LIBERO, 85.7% to 90.9% on LIBERO-plus, and 63.4% to 90.0% across four real-robot tasks without additional inference overhead. Trajectory analyses reveal that explicitly structuring trajectory priors serves most effectively as a learning inductive bias rather than a runtime constraint.
Authors: Ziyan Xiao, Yinghao Zhu, Wenting Zhang, Heaju Kim, Lequan Yu
Abstract: Multimodal large language models (MLLMs) are rapidly evolving with expanded context windows and stronger reasoning capabilities, enabling multi-chart understanding and multi-step inference. These abilities are increasingly important as MLLMs are adopted in complex agentic tasks. However, existing benchmarks largely emphasize single-chart perception, while simple chart-to-chart connections are insufficient to evaluate these capabilities. To capture multi-chart complexity while ensuring consistency and validity, we design a synthesis pipeline supported by latent graphs. Building on this pipeline, we introduce LongChart, a benchmark whose VQA sets contain an average of 6.5 images and 31.2 questions. We evaluate 10 state-of-the-art MLLMs and examine three factors that influence performance: reasoning patterns, auxiliary tools, and robustness to image perturbations. Our results show that MLLM accuracy decreases and varies substantially as computational complexity increases, highlighting directions for future research in multi-chart reasoning.
Authors: Ruiteng Zhao, Zhengshen Zhang, Yue Su, Wenshuo Wang, Jiahui Li, Zhiyuan Yang, Francis E. H. Tay, Marcelo H. Ang Jr., Haiyue Zhu
Abstract: World Action Models (WAMs) couple action generation with prediction of future states. Their effectiveness depends on whether future dynamics are modeled in a space that is both aligned with action generation and sufficiently geometry-aware to capture where and how actions change the scene. Existing WAMs typically satisfy only part of this requirement, relying on either perceptually heavy observation-space targets or auxiliary latent spaces that are not jointly structured for action relevance and geometry. We propose SG-WAM, a self-guided framework that learns geometry-aware action-conditioned dynamics directly in the policy-derived representation space. SG-WAM introduces learnable dynamics tokens and a Self-Guided World Predictor that forecasts their future latent states conditioned on intervening robot actions. Prediction targets are generated by an exponential moving average copy of the same policy backbone, providing stable supervision within the representation family used by the action expert. Geometric supervision further structures the policy image-token representations, providing spatially grounded context for the dynamics tokens and yielding a future-alignment space that is both action-relevant and geometry-aware. Latent future prediction, geometric grounding, and flow-matching action generation are jointly optimized end-to-end in a unified framework. Built on a 0.9B model without large-scale embodied pretraining, SG-WAM achieves 98.5% average success on LIBERO and 73% on LIBERO-Plus, while outperforming strong baselines in both in-distribution and out-of-distribution real-world evaluations.
Authors: Zeyu Ling, Xinyao Yu, Renye Yan, Jikang Cheng, Zhanke Wang, Qing Shuai, Changqing Zou
Abstract: General-purpose humanoid trackers can execute diverse references, but their zero-shot coverage depends on large embodied corpora that are costly to extend. Text-to-motion generators offer scalable supervision, yet models trained on human motion or retargeted data inherit a gap between kinematic plausibility and robot executability. Existing one-way pipelines fix either the generated corpus or the reward tracker. We introduce GenTrack, an online generator--tracker framework that alternates execution-grounded, group-relative generator alignment with tracker training on newly generated references; anchoring and rehearsal constrain drift. On Unitree G1, we evaluate GenTrack with ProtoMotions and SONIC backbones across three zero-shot tracking splits including public AMASS and LAFAN benchmarks, and a private out-of-distribution test set of 1,024 prompt-motion pairs in the wild. The online co-training strategy consistently produces generators that output more robot-executable motions with strong semantic alignment, and trackers with markedly broader zero-shot coverage and improved tracking accuracy, especially on out-of-distribution references. These results demonstrate that joint online post-training effectively narrows the executability gap between retargeted references and robot-native motion, advancing zero-shot humanoid control without additional data collection and beyond the limitations of a static reference pool.
Authors: Haoran Liao, Pengyue Wang, Shuoyu Chen, Kehan Cheng, Xuhang Chen, Yuhao Lin, Mu Lin, Zhizhao Liang, Xiaoyi Fan, Chengyi Xing, Dan Niu, Yi-Lin Wei, Wei-Shi Zheng
Abstract: Dynamic manipulation is a critical capability for robots operating in complex and dynamic environments, where robots must interact with objects that are moving or require rapid adjustments. However, learning models for dynamic manipulation tasks face two major challenges: (1) the combinatorial complexity of dynamic scenarios leads to substantial data requirements, and (2) rapid variations in dynamics require real-time and accurate policy execution. In this paper, we propose DynamicManip to address these challenges through an efficient data augmentation pipeline and a low-latency imitation policy. We first propose a static-to-dynamic augmentation pipeline that synthesizes diverse dynamic manipulation demonstrations from a single static demonstration. Second, we introduce a dynamic-aware adaptive policy that adjusts its inference frequency according to task dynamics, enabling responsive and effective dynamic manipulation. Third, we build a dynamic manipulation benchmark, which includes diverse dynamic tasks with an automatic evaluation system for scalable and consistent assessment. Extensive experiments in both simulation and the real world demonstrate that DynamicManip not only provides significant improvements in data efficiency but also achieves better performance in dynamic manipulation tasks, with a mean success rate 18.4 percentage points higher and policy-query latency 32.9% lower.
Authors: Alireza Furutanpey, Qiyang Zhang, Yujie Huang, Philipp Raith, Schahram Dustdar
Abstract: Earth-observation satellites capture more imagery than intermittent ground contacts can transmit. Onboard systems threshold a cloud detector, discard frames or tiles, and compress the survivors with a fixed codec. On expert-labeled imagery, these rules remove more than one-fifth of clear pixels, primarily through detector false positives. We train a neural codec with a clear-probability-weighted reconstruction loss, reallocating coded bytes from clouds to clear ground without requiring or transmitting a cloud map onboard. Each capture is encoded into a resumable base layer and a dependent refinement layer, while clear content is estimated from features produced by the encoder. At each contact, we causally rank arrived layers using estimated clear content, unfinished bytes, deadline slack, and aggregate deadline pressure. The scheduler serves base and computational deadlines, bounds stored residual bytes, and resumes interrupted packets. We evaluate the onboard-to-downlink pipeline using real entropy-coded bytes, orbit-derived interruptible contact capacities, and measured service time and energy on resource-constrained embedded accelerators. Clear-weighted codecs require up to 47.8\% fewer bytes than learned-compression baselines at matched clear-region quality. The optimized encoder consumes less time and energy than one pass of the cloud detector used by the frame-discard rules. Relative to fixed two-stage service on the same streams, our scheduler more than doubles deadline-full clear-content delivery for the interrupted combined cohort, reaches 83.6\% of a certified clairvoyant upper bound, and exceeds replayed reference orders in deadline-usable delivery.
Authors: Petr Somol, Ji\v{r}\'{\i} Grim
Abstract: Sequential subset search -- forward selection with floating backtracking and its descendants -- remains the quality reference in feature selection, but every member of the family sweeps the full pool of remaining candidate features at each step, which excludes it from very-high-dimensional problems; there, only individual-feature ranking remains practical, and it models feature interplay weakly or not at all. We introduce a budgeted sampled step operator pair that replaces the full sweeps by a fixed number of candidate evaluations per step. Candidates are drawn by temperature-controlled softmax sampling from dependency-aware per-feature statistics learned online from every criterion evaluation the search performs, guarded by a uniform exploration floor; per-step cost becomes independent of dimensionality. Substituting the operators turns any sequential method into its stochastic counterpart, defining the Stochastic Sequential Search (SSS) family; we study the stochastic counterpart of floating search, sSFFS. On 500-dimensional madelon, sSFFS retains at least 97% of the full-SFFS criterion value at every subset size at about a quarter of its evaluations, while uniform sampling at the same budget collapses on madelon's synergistic features. On 5,000-dimensional gisette, far beyond full-SFFS reach, sSFFS exceeds the saturated criterion level of DAF and BIF ranking at matched budgets; holdout validation shows that at 500 training samples the binding constraint beyond the sequential frontier becomes the criterion, not the search. On 10,105-dimensional reuters, under a trustworthy multinomial filter criterion, sSFFS dominates BIF and DAF on the search objective and on holdout accuracy at every subset size, in about two minutes of single-core evaluation work. A verified standalone implementation accompanies the paper.
Authors: Dingyi Kang, Dongming Jiang, Yi Li, Guanpeng Li, Bingzhe Li
Abstract: Interaction between users and LLM agents is increasingly multimodal: conversations interleave text with images, and a later question may target either. Yet most agent memories are designed around text, and even the few that support multimodal conversations still fail on vision-related questions. We trace this failure to an assumption behind the similarity search they rely on: in the index space, a query lies close to the relevant evidence that answers it. In multimodal settings, two gaps break it. By the modality gap, a query lies closer to memory content of its own modality than to evidence in another, even in a trained joint embedding space. By the similarity-relevance gap, the content most similar to a query is often not the evidence that answers it, most acutely when a query carries both text and image and its evidence resembles neither part alone. We present V-Mem, a multimodal agentic memory system that routes retrieval by the modality of the query and that of the target evidence, both recognized from the query alone. To cross the modality gap, V-Mem organizes the conversation into rounds and returns the target-modality content from the same round as the match, without comparing across modalities. To close the similarity-relevance gap, it searches with an LLM-generated anchor that sits closer to the relevant evidence than the query does: a hypothetical caption for a text-only query seeking an image, and an enriched search anchor, the query text plus relevant keywords extracted from the query image, when the evidence is reachable only by combining the two. On Mem-Gallery, V-Mem reaches an LLM-judge score of 0.82 versus 0.56 for the second best, with the largest margin on questions carrying an image (0.87, no baseline above 0.47); on LoCoMo it scores 0.69 versus 0.58.
Authors: Fabian Slonimczyk, Danila Karapsin
Abstract: The CLIP Q-score is a novel, safe, fully reproducible, and computationally efficient method for extracting objective product quality metrics from visual data using contrastive language-image pre-training. We introduce the technique and provide an extensive application to real estate data from an online platform ($\sim500,000$ images). Our open-source metric aligns with LLM assessments and proves to be a powerful predictor of housing market prices for both sales and rentals. We also show that a higher CLIP Q-store is associated with better liquidity (reduced time on the market), especially for properties on sale.
Authors: Chenchen Peng, Mixia Wu, Qijing Yan, Da Chen, Zhiqi Shen
Abstract: Detecting a change in a multivariate series answers only the first of two questions; the operational question is which coordinates changed. Existing answers are incomplete. Block-level procedures certify predefined groups of coordinates under an additive union bound, high-dimensional variable-selection methods return interpretable rankings without error guarantees, and the post-detection inference literature controls error along the time axis rather than across coordinates. We propose ARM (Attribution by Rank Maxima), a wrapper that accepts a changepoint located by an arbitrary detector and returns the set of coordinates certified to have changed, each carrying a location or scale type label. ARM scores each coordinate by a max-over-splits rank statistic. Because this statistic dominates the corresponding statistic at the estimated split, the resulting certificate is invariant to the manner, and to the accuracy, of the changepoint estimate. Three finite-sample guarantees follow from within-coordinate ranks alone: per-coordinate validity under any detector; exact family-wise error control through a Westfall--Young joint permutation that preserves cross-coordinate dependence, with a fully distribution-free Holm fallback; and false discovery rate control under arbitrary coordinate dependence in high dimensions through Benjamini--Yekutieli and e-BH. In simulations, naive per-coordinate testing at the estimated changepoint inflates its family-wise error beyond $0.66$ as the dimension grows, whereas ARM maintains the nominal level while retaining validity under heavy tails, power in high dimensions, and accurate type labels. On five financial series surrounding the 2008 collapse, ARM attributes a scale change to every asset class and excludes injected control coordinates.
Authors: Leyang Chen, Junyi Wu, Shaoqiu Zhang, Yulun Zhang
Abstract: Diffusion world models generate high-quality futures, but re- peated transformer evaluations make inference prohibitively slow. Existing caches reuse intermediate features, selectively update tokens, or reuse and extrapolate denoising outputs ac- cording to local drift or short native-space histories. These criteria can miss both approximation-induced latent transition defects that accumulate across skipped steps and phase- or condition-dependent changes in the direction of latent evo- lution. We propose WorldDynCache, a risk-controlled latent dynamics approximation framework with two core compo- nents. First, a lightweight latent-transition risk estimator tracks the accumulated future impact of approximation defects and calibrates its predictions against counterfactual defects ob- served at exact anchors. Second, a condition- and phase- aware lifted latent surrogate approximates latent evolution without extra transformer evaluations. On HunyuanVoyager- 13B and Aether-5B, WorldDynCache achieves 4.92 times and 2.15 times speedups, respectively, while attaining the best gen- eration quality among the compared caching methods across WorldScore, PSNR, SSIM, and LPIPS.
Authors: Lingwei Dang, Ziyan Qiu, Jiajia Cheng, Shishuo Shang, Zhenhao Zhang, Yufei Zhu, Qingxin Xiao, Pan Liu, Shenghui Huang, Yun Hao, Juntong Li, Qingyao Wu
Abstract: Existing indoor layout generators produce globally plausible layouts yet may retain local violations such as collisions, out-of-bounds placements, obstructed openings, and blocked circulation. Most prior work focuses on full-scene synthesis or scene-level optimization, with limited support for identifying responsible objects and locally repairing affected regions. We present Roomer, a reflective repair framework that casts these violations as sparse, object-grounded repair problems. Roomer encodes layouts as ``RoState'' and uses ``RoReview'' to bind measured violations to implicated objects. A geometry-conditioned vision-language model planner proposes a structured local edit, while a deterministic solver validates it and generates a finite set of candidate edits when needed. Each candidate is committed only if full-scene verification confirms that it resolves the target violation without new hard violations or broken protected constraints. We train the planner on Roomer-CC, a controlled-corruption dataset that pairs faulty layouts with object-grounded violation evidence and known-feasible inverse StatePatches. Since existing benchmarks rarely assess whether physically valid layouts are usable, we introduce Roomer-Eval to assess distributional quality, physical validity, and practical usability. Experiments show that Roomer repairs residual violations while preserving valid regions, improves physical validity and usability, and transfers across external generators.
Authors: Zhaozhi Luo, Janne Heiskanen, Ilja Vuorinne, Ian Ocholla, Shiqi Zhang, Saana J\"arvinen, Xinyu Wang, Yanfei Zhong, Petri Pellikka
Abstract: The beekeeping sector holds significant potential for livelihood diversification among the agropastoral communities in Kenya. Melliferous tree species play a critical role by providing essential nectar sources for bees. However, limited knowledge of their precise spatial distributions constrains the full development of beekeeping. One-class classification (OCC) offers a practical solution for detecting single target species without requiring extensive labeled data from other classes. Although existing OCC methods perform well in trained domains, the generalization capability to unseen domains remains limited due to domain shift. To address these challenges, this study proposes a hyperspectral unsupervised domain adaptation OCC framework (HyUDA-One) for tree species mapping using airborne hyperspectral imagery and laser scanning data. The spatial-spectral regularized pseudo-positive learning was designed to mitigate domain shift and improve model generalizability. The effectiveness of HyUDA-One was demonstrated by mapping three key melliferous tree species in two savanna landscapes in southern Kenya. The results show that HyUDA-One significantly improves performance in unlabeled domains. The F1-scores of 0.788, 0.845, and 0.768 were achieved for Senegalia mellifera, Vachellia tortilis, and Commiphora africana in the trained domain, respectively. In the untrained domain, the F1-scores of Senegalia mellifera and Vachellia tortilis were 0.756 and 0.884, respectively. The distribution maps revealed the spatial patterns of these melliferous tree species and the nectar source availability, offering an important reference for sustainable beekeeping development in savanna landscapes. Furthermore, the proposed framework can potentially be extended to other mapping applications, such as invasive species detection.
Authors: Andrea Brigliadori, Leevi Kerkela, Hui Zhang
Abstract: Brain tissue microstructure estimation with machine learning provides higher computational efficiency than conventional fitting. However, machine learning still presents important limitations that hamper its clinical utility. Specifically, current models typically lack generalisation across diffusion MRI acquisition protocols and require retraining whenever b-vectors or b-values change. Moreover, the recent machine learning methods that were developed to address protocol generalisation lack rotational equivariance. Particularly suitable for dMRI parameter estimation is a geometric deep learning model known as spherical convolutional neural network (SCNN), which guarantees rotational equivariance and b-vector generalisation. However, this architecture currently does not account for b-values. Therefore, obtaining a model that combines protocol generalisation and rotational equivariance remains an open challenge. In this paper, we directly address this issue by incorporating explicit b-value dependence into an SCNN architecture via a hypernetwork. This new approach is illustrated using NODDI as an example forward model for estimating brain tissue microstructure. To evaluate b-value generalisation, the original and newly proposed SCNN architectures are trained on synthetic data and tested on both synthetic and real data across different b-value pairs. Results demonstrate that the proposed method achieves reduced RMSE and bias on synthetic data, as well as higher agreement with conventional NODDI fitting on real data, indicating improved robustness to unseen b-values and a reduced need for retraining. By combining generalisation across b-values with generalisation across b-vectors and rotational equivariance, the proposed framework enhances the applicability of deep learning to clinical diffusion MRI parameter estimation. Code available at https://github.com/aerdnairo/arXiv\_generalisedSCNN.
Authors: Wei Jia, Zhicong Lu, Yu Chen, Xiang Wang, Shuai Li, Wenqian Lv, Jiayue Cao, Huaxing liu
Abstract: Large vision-language models (LVLMs) have achieved substantial performance gains in Video Temporal Grounding (VTG) through reinforcement learning (RL). However, existing methods primarily rely on outcome correctness rewards that evaluate only the final predicted intervals, leaving boundary-related visual evidence and its correspondence with timestamp predictions insufficiently constrained. In this paper, we delve into timestamp prediction and its underlying boundary-level visual evidence, showing prevalent misalignment between visual evidence and predicted timestamps across widely used benchmarks. To address this issue, we propose Competence-Aware Visual Boundary Evidence Alignment (CAVE), which augments localization optimization with boundary-specific visual evidence rewards to mitigate evidence-timestamp misalignment. Specifically, to explicitly represent the boundary-specific visual evidence, CAVE introduces boundary-specific evidence tokens and initializes their structured generation and distinct boundary semantics through a lightweight supervised warm-up. During RL, the visual boundary evidence alignment reward reinforces the visual attention of special evidence tokens within the ground-truth boundaries, thereby promoting alignment between visual evidence and temporal boundaries. Moreover, performance-aware gating for evidence supervision is designed to adaptively retain evidence guidance for poorly localized groups while reducing it once localization becomes sufficiently accurate to avoid over-constraining fine-grained boundary refinement. Extensive experiments on several public VTG benchmarks demonstrate the effectiveness of our method.
Authors: Minnan Pei, Gang Li, Zeyu Zhu, Siting Wang, Junwen Si, Zhuoran Song, Yu Feng, Fangxin Liu, Xiaoyao Liang, Jian Cheng
Abstract: 3D Gaussian Splatting (3DGS) has emerged as a leading technique for real-time novel view synthesis, yet existing 3DGS accelerators suffer from poor architectural scalability: increasing the number of PEs leads to marginal performance improvement during rendering. We identify that the root cause is the tightly coupled ``checking-while-blending'' dataflow, which exacerbates PE underutilization caused by spatial redundancy from irregular Gaussian coverage and temporal redundancy from asynchronous pixel-wise termination under parallel execution. To address this issue, we propose DeGS, a scalable architecture for efficient 3DGS inference. To systematically eliminate the redundancies inherent in rendering, DeGS exploits a decoupled dataflow, restructuring the coupled $\alpha$-checking, transmittance checking, and $\alpha$-blending of the standard rendering process into consecutive workload parsing, reorganization, and blending stages. This allows the fragmented, length-variable, and temporal-dependent workloads to be reorganized into compact, conflict-free, and dense workloads prior to blending, thereby significantly improving PE utilization during parallel blending. Implemented in 28 nm technology, DeGS achieves 2.36$\times$--7.25$\times$ throughput, 1.82$\times$--6.02$\times$ end-to-end speedup, and 1.59$\times$--4.42$\times$ energy efficiency over state-of-the-art 3DGS accelerators (GSCore, GBU, GCC) across diverse scenes and resolutions (720p to 8K). Moreover, scaling from 16 to 1024 PEs, DeGS maintains over 80\% PE utilization at high resolutions, significantly outperforming existing accelerators.
Authors: Haonan Chen, Chu Li, Zhicheng Wang, Yuanwei Liu, Yuanjiang Wang, Shaohua Jiang, Zhicheng Dou
Abstract: Multimodal representation learning is a cornerstone of modern AI. By encoding multimodal queries and targets into vectors, it powers industrial search and recommendation and underpins modern agents. Real-world platforms with complex modalities and massive-scale content, such as Douyin, Xiaohongshu, and YouTube, demand both efficiency under billion-scale indexing and fine-grained discrimination for hard matching. Existing MLLM embedding models rarely satisfy both. Contrastive models are efficient but rely on pair-level supervision too coarse for fine-grained distinctions, while CoT-based models improve discrimination through explicit generation impractical to serve online. We present Douyin Multimodal Embedding (DME), a model trained in two stages to combine both strengths. Stage 1 performs large-scale contrastive pre-training that establishes a unified multimodal embedding space with broad modality and task coverage. Stage 2 supplements semantic sufficiency, the property that an embedding is grounded in retrieval-relevant evidence and preserves fine-grained counterpart-side semantics, via two mechanisms. Evidence-Grounded Typed Latent Reasoning organizes retrieval evidence through hidden-space latent reasoning, and Cross-Conditional Reconstruction enforces counterpart-side semantics through cross-directional autoregressive reconstruction. Both act only during training and add only marginal query-side overhead, so DME serves as efficiently as a standard contrastive encoder. On MMEB-v2, DME reaches state-of-the-art results at comparable scales for its 2B and 9B variants (74.8 and 78.4), with especially strong video and visual-document tasks. In production, DME delivers a 2.92% relative gain on Douyin's in-house offline evaluation set, is deployed across Douyin scenarios such as generative, image, and AI search, and yields a 0.1% Lifetime (LT) gain in online A/B testing on Douyin search.
Authors: Onur Ali Zeybekoglu, David Tilly, Orcun Goksel
Abstract: Deformable image registration (DIR) is a core problem in medical image analysis; but, unlike labeling decision problems such as classification and segmentation, registration is a problem class that involves stringent physical constraints. Although deep learning methods have made faster registration possible, the resulting models are often difficult to interpret compared to hand-crafted methods with explicit objectives and interpretable physical meaning. In this work, we show that an analytical method can still yield competitive and superior results to deep learning in a common deformable registration task. We study pTVreg as a parametric total variation based registration in that context. Observing its different implementations to perform at various degrees, we introduce here an accessible implementation of this method, together with a Bayesian optimization framework that automatically sets self-parameters for any DIR task from a set of sample examples. Experiments on Lung250M-4B show that our proposed implementation achieves state-of-the-art results in this benchmark, substantially superior to existing deep learning solutions and other pTVreg variants as baselines. The source code will be made publicly available at https://github.com/oazeybekoglu/ptvreg-python .
Authors: Daniel Fusaro, Simone Mosco, Wanmeng Li, Alberto Pretto
Abstract: Traversability analysis is a fundamental capability for autonomous mobile robots operating in unstructured environments. While modern machine learning approaches such as deep neural networks and gradient-boosted trees achieve strong predictive performance, they lack interpretability and provide limited insight into the underlying terrain-robot interaction dynamics. In this paper, we propose TravKAN, a Kolmogorov-Arnold Network-based framework for fast, scalable, and interpretable traversability estimation. TravKAN represents multivariate decision functions through compositions of learnable univariate functions, enabling compact architectures and symbolic extraction of analytic expressions after training. In addition, we introduce a novel set of handcrafted features derived from the reflectivity channel of LiDAR sensors. To the best of our knowledge, reflectivity has not been systematically exploited for handcrafted traversability descriptors, despite its potential to capture material and surface properties complementary to geometric cues. We evaluate TravKAN on public, real-world urban and off-road datasets and compare it against strong baselines. TravKAN achieves strong performance across all metrics, outperforming conventional deep models and approaching the performance of XGBoost. TravKAN-Lite, i.e., TravKAN's symbolic representation, reveals meaningful nonlinear feature interactions and provides a compact, deployment-friendly, and fast analytic model. Ablation studies further show the robustness of our method to architectural variations and quantify the contribution of the proposed reflectivity-based features. These properties make TravKAN attractive for robotic systems requiring transparency, real-time computational efficiency, and interpretability in safety-critical decision-making.
Authors: Michael Farmer
Abstract: Can scientific abduction occur without continuous sensorimotor embodiment? Recent arguments in AI and philosophy of science hold that genuine hypothesis generation requires an agent continuously coupled to the physical world. We defend a narrower claim: online embodiment is not necessary for every abductive scientific act. Our focus is identity abduction: the inference that two independently developed structures are one object under an explicit correspondence, reached through representational grounding rather than bodily interaction. An agent may acquire new inferential affordances not through physical interaction but through transformations into representations that expose latent invariants. Scientific diagrams are a practical substrate because they embody independently evolved conventions that partially canonicalize symmetry, topology, and operator structure across disciplines - a property we develop as convention space, which answers a hard retrieval problem: finding mathematically related work when two fields share no discriminating vocabulary. We operationalize the mechanism as an architecture, the Abduction Loop: representation generation, motif extraction, convention-space canonicalization, cross-domain retrieval, identity-hypothesis generation, and adversarial verification, with abstention as the designed default. A documented episode, in which a multimodal model given a figure of a gravitational-memory transport model generated and then verified the hypothesis that its central differential complex is equivalent to the spherical Kaiser-Squires mass-mapping complex of weak-lensing cosmology, serves as a motivating possibility witness from which the architecture is abstracted, not as evidence of general capability. We close with a falsifiable evaluation program, the DAB-30 benchmark. The contribution is a mechanistic proposal, an architecture, and a test program.
Authors: Luc Trudeau, Maria G. Martini
Abstract: Efficient and perceptually meaningful quality assessment is a fundamental requirement for image and video processing, compression, and streaming systems. This article shows that, in the context of Discrete Cosine Transform ( DCT)-based compressed images, Structural Similarity Index ( SSIM ) can be approximated from global Peak Signal to Noise Ratio (PSNR) or Mean Square Error ( MSE) using local statistics derived only from the reference image. While prior work assumes access to local MSE, we propose two approaches to approximate local MSE by redistributing the global MSE using variance or standard-deviation-based weighting. Experiments on the Kodak and Xiph Subset1 datasets across a range of JPEG quality levels demonstrate that both approaches provide accurate and robust SSIM approximations, substantially outperforming the global MSE baseline. The proposed framework is designed to extend naturally to video, where reference-derived statistics can be amortized across multiple encodes of the same content.
Authors: Xiang An, Xuhan Zhu, Yang Xiao, Lan Wu, Ming Zhang, Yuan Gao, Bin Qin, Debing Zhang, Ying Fu, Jiankang Deng
Abstract: Training face recognition models with millions of identities is challenging because classifier storage, logit memory, and computation grow linearly with the number of classes, eventually making full softmax impractical even when the backbone itself fits comfortably in memory. We present Partial FC (PFC), a scalable approximation to large-class softmax that preserves every positive class center while activating only a sampled subset of negative centers in each mini-batch. This asymmetric treatment retains every target term while avoiding exhaustive interaction with millions of mostly uninformative negatives. Our distributed implementation partitions the classifier across GPUs and samples within each owned shard, so sampling reduces local matrix multiplication and logit storage while sharding eliminates class-gradient synchronization across workers. Together, these properties reduce GPU-resident classifier memory, computation, and class-dependent communication without feature-based hard-negative retrieval. End-to-end system benchmarks demonstrate efficient scaling to massive class spaces, including 64 million classes on a single eight-GPU machine. Across large-scale face-recognition datasets, moderate sampling rates maintain competitive recognition accuracy while substantially improving training efficiency. Our best PFC configurations achieve 97.2\% TAR on IJB-C at FAR $=10^{-4}$ and 94.0\% TAR on ICCV21-MFR at FAR $=10^{-6}$. Beyond clean training data, PFC is robust to inter-class conflicts, label noise, and long-tailed identity distributions: under 40\% label noise, PFC with conflict filtering raises ICCV21-MFR TAR from 43.9\% to 80.2\%, while on long-tailed data PFC improves TAR from 87.4\% to 92.0\%. These results establish positive-preserving negative sampling as an effective foundation for scalable, accurate, and robust identity classification.
Authors: Ancheng Lin, Tianqing Su, Zuo Yuan, Quanke Su, Samuel S. Mao, Yusheng Xiang
Abstract: Jointly recovering explicit surface geometry and high-quality appearance from multi-view images remains challenging. This capability is essential for maintaining high-fidelity real-to-sim environments for embodied intelligence, where local changes should be incorporated without complete reconstruction. Existing neural surface reconstruction and 3DGS-to-mesh pipelines often learn geometry indirectly or separate geometry construction from appearance modeling. This separation introduces optimization redundancy and makes local geometry or appearance updates expensive. We propose an end-to-end mesh-Gaussian scene representation that binds 3D Gaussians to mesh faces and uses differentiable 3DGS rendering for photometric supervision. This design provides a direct information pathway for jointly learning explicit geometry and renderable appearance. Experiments on indoor and outdoor scenes demonstrate improved efficiency and rendering quality while preserving high-quality surface reconstruction. The explicit mesh also enables mesh-based manipulation, and the coupled representation adapts efficiently to local scene modifications. These properties support scalable visual scene modeling and the efficient maintenance of real-to-sim environments for embodied-agent training and evaluation.
Authors: Asma Khan, Tushar Kataria, Janmesh Ukey, Shireen Y. Elhabian
Abstract: Statistical Shape Models (SSMs) excel at identifying population level anatomical variations, which is at the core of various clinical and biomedical applications, including morphology-based diagnostics and surgical planning. However, the effectiveness of SSMs is often constrained by the necessity for expert-driven manual segmentation, a time-intensive and expensive process that restricts their broader utility. While deep learning approaches offer a potential workaround by directly estimating SSMs from unsegmented images, they merely shift the burden. Although these models do not require segmentation during deployment, they still fail to address the challenge of acquiring the manual annotations needed for training, particularly in resource-limited settings. Semi-supervised models for anatomy segmentation present a logical solution to the annotation burden. However, the lack of established guidelines leaves end-users uncertain about the actual effectiveness of these approaches for the downstream task of constructing SSMs. In this study, we bridge this gap by systematically evaluating semi-supervised methods as viable alternatives to manual segmentation. By applying these methods under low-annotation settings and utilizing the predicted segmentations for SSM generation, we establish a comprehensive new performance benchmark. Our findings reveal a clear divide in performance: while certain methods yield noisy segmentations that degrade SSM quality, others accurately capture the population's modes of variation comparable to those obtained from manual-segmentation SSMs despite a 60-80% reduction in manual annotation requirements.
Authors: Thanh-Dung Le, Vu Nguyen Ha, Ti Ti Nguyen, Duc-Dung Tran, Hung Nguyen-Kha, Luis M. Garces-Socarras, Juan Carlos Merlano-Duncan, Symeon Chatzinotas
Abstract: Remote sensing (RS) image classification is central to Earth observation, but onboard deployment requires models that are accurate, efficient, and robust to sensor and transmission degradation. Following a train-on-ground, infer-onboard workflow, we evaluate 14 backbones, including CNNs, ResNets, compact Transformers trained from scratch, and pre-trained Vision Transformers, on EuroSAT and PatternNet. We assess clean-data performance, computational cost, power consumption, and robustness to Gaussian noise, motion blur, and an end-to-end DVB-S2(X) transmission chain with channel impairments and JPEG compression. Pre-trained Vision Transformers generally outperform models trained from scratch while providing better efficiency and corruption resilience. MobileViTV2 achieves the highest clean EuroSAT accuracy at 99.09%, whereas EfficientViT-M2 provides the strongest overall trade-off. It attains 98.76% accuracy, precision, and recall on EuroSAT and 99.52% accuracy on PatternNet, with 203.53 MFLOPs, a 38.19 MB footprint, and the best overall robustness score of 0.79. It also degrades most gracefully under transmission loss and consumes 63.35% less power than MobileViTV2 and 73.33% less than Swin Transformer. These results identify EfficientViT-M2 as a strong backbone for reliable, energy-efficient onboard RS image classification. Code for data augmentation, corruption generation, training, and inference is publicly available.
Authors: Guofeng Mei, Qinfeng Xiao, Bin Ren, Luigi Riz, Juan Liu, Xiaoshui Huang, Xu Zheng, Nicu Sebe, Ming-Hsuan Yang, Fabio Poiesi
Abstract: Vision-language models, such as CLIP, encode rich semantic knowledge through large-scale image-text pretraining. Reusing these models for 3D understanding is highly desirable, because 3D-text pairs and dense point-level annotations are far scarcer and more difficult to obtain than their 2D counterparts. However, CLIP is trained on regular 2D image patches, whereas point clouds are unordered, sparse, and irregular. The primary challenge lies in constructing 3D token sequences whose geometry, position, and local relations can be interpreted by a frozen, pretrained vision transformer. Moreover, coordinate-scale variations across heterogeneous 3D domains make unified training particularly challenging. To bridge these gaps, we propose a parameter-efficient framework that learns a scale-normalized 3D tokenizer, enabling the CLIP visual encoder to be reused for point-cloud understanding. Our tokenizer, named UTok3D, estimates an input-adaptive geometric scale to calibrate sparse voxelization, coordinate normalization, token-center computation, and 3D positional encoding, and serializes the resulting tokens by Hilbert ordering. This shared geometric convention makes joint training feasible across point clouds with substantially different metric scales, including object-level shapes, indoor scenes, and outdoor LiDAR scans. The tokenizer is trained without 3D annotations through self-supervised cross-modal distillation from foundation-model features extracted from multi-view images, combining local superpoint alignment with the proposed Sinkhorn Ranked Contrastive distillation. Experiments on ShapeNetPart, ScanNetV2, S3DIS, SemanticKITTI, and nuScenes show that, when equipped with a scale-consistent and CLIP-interpretable token interface, a frozen CLIP visual backbone can be effectively reused for annotation-free 3D segmentation.
Authors: Rachid Zeghlache, Ikram Brahim, Pierre-Henri Conze, Mathieu Lamard, Mohammed El Amine Lazouni, Zineb Aziza Elaouaber, Leila Ryma Lazouni, Christopher Nielsen, Ahmad O. Ahsan, Matthias Wilms, Nils D. Forkert, Lovre Antonio Budimir, Ivana Matovinovi\'c, Donik Vr\v{s}nak, Sven Lon\v{c}ari\'c, Philippe Zhang, Weili Jiang, Yihao Li, Yiding Hao, Markus Frohmann, Patrick Binder, Marcel Huber, Taha Emre, Teresa Finisterra Ara\'ujo, Marzieh Oghbaie, Hrvoje Bogunovi\'c, Amerens A. Bekkers, Nina M. van Liebergen, Hugo J. Kuijf, Abdul Qayyum, Moona Mazher, Steven A. Niederer, Alberto J. Beltr\'an-Carrero, Juan J. G\'omez-Valverde, Javier Torresano-Rodr\'iquez, \'Alvaro Caballero-Sastre, Mar\'ia J. Ledesma Carbayo, Yosuke Yamagishi, Yi Ding, Robin Peretzke, Alexandra Ertl, Maximilian Fischer, Jessica K\"achele, Sofiane Zehar, Karim Boukli Hacene, Thomas Monfort, B\'eatrice Cochener, Mostafa El Habib Daho, Anas-Alexis Benyoussef, Gwenol\'e Quellec
Abstract: The MARIO challenge, held at MICCAI 2024, focused on advancing the automated detection and monitoring of age-related macular degeneration (AMD) through the analysis of optical coherence tomography (OCT) images. Designed to evaluate algorithmic performance in detecting neovascular activity changes within AMD, the challenge incorporated unique multi-modal datasets. The primary dataset, sourced from Brest, France, was used by participating teams to train and test their models. The final ranking was determined based on performance on this dataset. An auxiliary dataset from Algeria was used post-challenge to evaluate population and device shifts from submitted solutions. Two tasks were involved in the MARIO challenge. The first one was the classification of evolution between two consecutive 2D OCT B-scans. The second one was the prediction of future AMD evolution over three months for patients undergoing anti-vascular endothelial growth factor (VEGF) therapy. Thirty-five teams participated, with the top 12 finalists presenting their methods. This paper outlines the challenge's structure, tasks, data characteristics, and winning methodologies, setting a benchmark for AMD monitoring using OCT, infrared imaging, and clinical data (such as the number of visits, age, gender, etc.). The results of this challenge indicate that artificial intelligence (AI) performs as well as a physician in measuring AMD progression (Task 1) but is not yet able of predicting future evolution (Task 2).
Authors: Yuxiang Wang, Xuecheng Bai, Chuanzhi Xu, Ying Zhou, Weidong Cai
Abstract: Small-object detection in Unmanned Aerial Vehicle (UAV) imagery requires preserving weak local evidence while using broader context to separate tiny foreground targets from cluttered backgrounds. Existing multi-scale fusion methods improve feature aggregation, but they often add computation or blur fine details during repeated cross-scale fusion. The central challenge is to balance low-SNR target preservation, clutter suppression, and efficient cross-scale context exchange. To address this challenge, we propose the Multi-scale Global-detail Feature Integration Strategy (MGDFIS), a neck-level feature-fusion strategy that couples global context exchange, local-detail recovery, and pixel-level foreground-background recalibration. MGDFIS integrates three coordinated modules: FusionLock-TSS Attention for stabilizing spectral-spatial responses, Global-detail Integration for combining long-range mixing with local detail capture, and Dynamic Pixel Attention for reweighting compact foreground regions. On the controlled VisDrone setting, YOLO26m + MGDFIS improves AP50:95 from 25.7 to 30.2 and AP50 from 37.2 to 44.2 over the YOLO26m baseline, with 96.1 GFLOPs. Additional dataset-specific evaluations report 38.9 AP50 and 21.9 AP50:95 on UAVDT and 97.4 AP50 on CARPK. The code is available at: https://github.com/Bai-Xuecheng/MGDFIS.
Authors: Varun Belagali, Pierre Marza, Srikar Yellapragada, Zilinghan Li, Tarak Nath Nandi, Ravi K Madduri, Joel Saltz, Stergios Christodoulidis, Maria Vakalopoulou, Dimitris Samaras
Abstract: Cross-view masked autoencoding has emerged as a powerful pretext task for learning dense correspondences, which are essential for applications such as video label propagation. The cross-view pretext task is modeled with a masked autoencoder, where a masked target view is reconstructed from an anchor view. However, acquiring effective training data remains a challenge - collecting diverse video datasets is costly, while simple image crops lack the necessary pose variations, underperforming video-based methods. This paper introduces CDG-MAE, a novel MAE-based self-supervised method that uses diverse synthetic views generated from static images via an image-conditioned diffusion model. We present a quantitative method to evaluate the local and global consistency of the generated views to choose the right diffusion model for cross-view self-supervised pretraining. These generated views exhibit substantial changes in pose and perspective, providing a rich training signal that overcomes the limitations of video and crop-based anchors. Furthermore, we enhance the standard single-anchor MAE setting to a multi-anchor masking strategy to increase the difficulty of the pretext task. CDG-MAE substantially narrows the gap to video-based MAE methods, while maintaining the data advantages of image-only MAEs.
Authors: Yanxing Liang, Yinghui Wang, Wei Li, Tao Yan, Jiaxing Shen
Abstract: We propose CLONE, a Continuous Latent Optimization framework for Normal Estimation via 3D Gaussian splatting. The core idea is to construct an image-geometry-image consistency loop that unifies explicit geometric representation with differentiable rendering, thereby enabling weakly supervised learning without normal ground truth. Specifically, CLONE comprises four components. First, by introducing a differentiable light interaction model with a learnable modulation kernel, we perform a unified reparameterization of the 3DGS parameter space, establishing an explicit and stable mapping between 3DGS geometric parameters and surface normals and turning the photometric loss into an internal supervision signal. Second, the conditional single-step deterministic refinement network integrates denoising architectures with differentiable reprojection constraints to refine the initial normals, thereby adaptively recovering the high-frequency details erased by the inherently smooth Gaussian primitives. Third, the cross-domain gating fusion mechanism adaptively combines the two complementary normal estimates while imposing multi-view reprojection consistency and implicit geometric regularization, reconciling the geometrically consistent yet over-smooth 3DGS estimate with the detailed yet potentially geometry-inconsistent refinement. Finally, all components are jointly optimized under a unified photometric reprojection objective with geometric consistency regularizations in a fully differentiable pathway, and the directional regularization aligns the learnable principal directions with the geometric normals, achieving an end-to-end optimization closed loop without relying on external normal labels.
Authors: Kaile Du, Zihan Ye, Junzhou Xie, Yixi Shen, Yuyang Li, Fuyuan Hu, Ling Shao, Guangcan Liu, Joost van de Weijer, Fan Lyu
Abstract: Multi-label class-incremental learning (MLCIL) continuously expands the label space while recognizing multiple co-occurring categories, making catastrophic forgetting a central challenge. Recent class-incremental learning methods have increasingly adopted CLIP as their backbone. However, we find that applying CLIP to MLCIL exhibits two critical issues: entanglement of class-specific cues in shared visual representations and high false-positive rates (FPR) under task-level partial labeling. We propose DeCLIP, a replay-free and parameter-efficient framework for CLIP-based MLCIL. DeCLIP uses Decoupled Prompting to learn class-specific positive and negative prompts in both visual and textual modalities, enabling class-conditioned vision-language matching and reducing representation entanglement. Only new-category prompts are optimized, previous prompts remain unchanged, preserving prior knowledge and mitigating catastrophic forgetting without replay. DeCLIP further incorporates Adaptive Similarity Tempering, an inference-time strategy that adapts similarity-tempering strength to the incremental configuration, suppressing false positives without specific tuning. Experiments on MS-COCO, PASCAL VOC, and the real-world NUS-WIDEseq benchmark demonstrate consistent improvements over prior methods with a few trainable parameters.
Authors: Yifan Liu, Zhiyuan Min, Zhenwei Wang, Junta Wu, Tengfei Wang, Yixuan Yuan, Yawei Luo, Chunchao Guo
Abstract: We present WorldMirror, a unified feed-forward model for comprehensive 3D geometric prediction tasks. Unlike existing methods constrained to image-only inputs or customized for a specific task, our framework flexibly integrates diverse geometric priors, including camera poses, intrinsics, and depth maps, while simultaneously generating multiple 3D representations: dense point clouds, multi-view depth maps, camera parameters, surface normals, and 3D Gaussians. Remarkably, prior injection yields universal gains across all tasks, suggesting that input flexibility and multi-task prediction are mutually reinforcing. WorldMirror achieves state-of-the-art performance across diverse benchmarks from camera, point map, depth, and surface normal estimation to novel view synthesis, while maintaining the efficiency of feed-forward inference.
Authors: Minxing Luo, Linlong Fan, Wang Qiushi, Ge Wu, Yiyan Luo, Yuhang Yu, Jinwei Chen, Yaxing Wang, Qingnan Fan, Jian Yang
Abstract: Current image super-resolution methods show strong performance on natural images but distort text, creating a fundamental trade-off between image quality and textual readability. To address this, we introduce TIGER (Text-Image Guided supEr-Resolution), a novel two-stage framework that breaks this trade-off through a "text-first, image-later" paradigm. TIGER explicitly decouples glyph restoration from image enhancement: it first reconstructs precise text structures and uses them to guide full-image super-resolution. This ensures high fidelity and readability. To support comprehensive training and evaluation, we present the UZ-ST (UltraZoom-Scene Text) dataset, the first Chinese scene text dataset with extreme zoom. Extensive experiments show TIGER achieves state-of-the-art performance, enhancing readability and image quality.
Authors: Yang Zhang, Qianyu Zhou, Farhad Imani, Jiong Tang
Abstract: Wind turbine blades operate in harsh environments, making timely damage detection essential for preventing failures and optimizing maintenance. Drone-based inspection and deep learning are promising, but typically depend on large, labeled datasets, which limit their ability to detect rare or evolving damage types. To address this, we propose a zero-shot-oriented inspection framework that integrates Retrieval-Augmented Generation (RAG) with Vision-Language Models (VLM). A multimodal knowledge base is constructed, comprising technical documentation, representative reference images, and domain-specific guidelines. A hybrid text-image retriever with keyword-aware reranking assembles the most relevant context to condition the VLM at inference, injecting domain knowledge without task-specific training. We evaluate the framework on 30 labeled blade images covering diverse damage categories. Although the dataset is small due to the difficulty of acquiring verified blade imagery, it covers multiple representative defect types. On this test set, the RAG-grounded VLM correctly classified all samples, whereas the same VLM without retrieval performed worse in both accuracy and precision. We further compare against open-vocabulary baselines and incorporate uncertainty Clopper-Pearson confidence intervals to account for the small-sample setting. Ablation studies indicate that the key advantage of the framework lies in explainability and generalizability: retrieved references ground the reasoning process and enable the detection of previously unseen defects by leveraging domain knowledge rather than relying solely on visual cues. This research contributes a data-efficient solution for industrial inspection that reduces dependence on extensive labeled datasets.
Authors: Forouzan Fallah, Wenwen Li, Chia-Yu Hsu, Hyunho Lee, Anna Liljedahl, Yezhou Yang
Abstract: High spatial resolution satellite imagery is critical for monitoring fine-scale Earth surface processes, but is often limited by cost and revisit time. This work studies cross-sensor super-resolution (SR) to reduce this gap by translating 10 m Sentinel-2 imagery into 2 m Maxar-like imagery in a data-scarce, domain-shifted setting, with a focus on rare geomorphic features such as retrogressive thaw slumps (RTS). We propose RareFlow, a semantic-guided generative AI framework for cross-sensor super-resolution based on a flow-matching formulation, designed to produce visually plausible and physically reliable high-resolution images. RareFlow uses dual conditioning to guide the generation process: (1) a gated ControlNet that preserves scene geometry from low-resolution (LR) input, and (2) text-based semantic guidance that injects contextual information when the target phenomenon is rare. To ensure high-fidelity outputs, we introduce a multifaceted loss function that anchors the output to the high-resolution (HR) ground truth by jointly enforcing frequency alignment, perceptual similarity, and color consistency. RareFlow's performance is systematically evaluated on a newly curated benchmark of multi-sensor satellite imagery for rare Earth feature detection, and its generalizability is demonstrated on two public remote sensing benchmarks, SEN2NAIP and BreizhSR. Human evaluation with domain experts is also conducted to further verify RareFlow's effectiveness in generating high-fidelity super-resolved images for scientific analysis.
Authors: Sichen Guo, Wenjie Li, Yuanyang Liu, Guangwei Gao, Jian Yang, Chia-Wen Lin
Abstract: Recently, Mamba-based super-resolution (SR) methods have demonstrated the ability to capture global receptive fields with linear complexity, addressing the quadratic computational cost of Transformer-based SR approaches. However, existing Mamba-based methods lack fine-grained transitions across different modeling scales, which limits the efficiency of feature representation. In this paper, we propose T-PMambaSR, a lightweight SR framework that integrates window-based self-attention with Progressive Mamba. By enabling interactions among receptive fields of different scales, our method establishes a fine-grained modeling paradigm that progressively enhances feature representation without introducing additional computational cost. Furthermore, we introduce an Adaptive High-Frequency Refinement Module (AHFRM) to recover high-frequency details lost during Transformer and Mamba processing. Extensive experiments demonstrate that T-PMambaSR progressively enhances the model's receptive field and expressiveness, achieving competitive performance with recent Transformer- or Mamba-based methods while incurring lower computational cost. Codes will be available at https://github.com/IVIPLab/T-PMambaSR.
Authors: Ege Ozguroglu, Junbang Liang, Ruoshi Liu, Mia Chiquier, Michael DeTienne, Wesley Wei Qian, Alexandra Horowitz, Andrew Owens, Carl Vondrick
Abstract: While olfaction is central to how animals perceive the world, this rich chemical sensory modality remains largely inaccessible to machines. One key bottleneck is the lack of diverse, multimodal olfactory training data collected in natural settings. We present New York Smells, a large dataset of paired image and olfactory signals captured ``in the wild.'' Our dataset contains 7,000 smell-image pairs from 3,500 distinct objects across indoor and outdoor environments, with approximately 70$\times$ more objects than existing olfactory datasets. Our benchmark has three tasks: cross-modal smell-to-image retrieval, recognizing scenes, objects, and materials from smell alone, and fine-grained discrimination between grass species. Through experiments on our dataset, we find that visual data enables cross-modal olfactory representation learning, and that our learned olfactory representations outperform widely-used hand-crafted features.
Authors: Zhoujie Fu, Xianfang Zeng, Jinghong Lan, Xinyao Liao, Cheng Chen, Junyi Chen, Jiacheng Wei, Wei Cheng, Shiyu Liu, Yunuo Chen, Gang Yu, Guosheng Lin
Abstract: Pre-trained video models learn powerful priors for generating high-quality, temporally coherent content. While these models excel at temporal coherence, their dynamics are often constrained by the continuous nature of their training data. We hypothesize that by injecting the rich and unconstrained content diversity from image data into this coherent temporal framework, we can generate image sets that feature both natural transitions and a far more expansive dynamic range. To this end, we introduce iMontage, a unified framework designed to repurpose a powerful video model into an all-in-one image generator. The framework consumes and produces variable-length image sets, unifying a wide array of image generation and editing tasks. To achieve this, we propose an elegant and minimally invasive adaptation strategy, complemented by a tailored data curation process and training paradigm. This approach allows the model to acquire broad image manipulation capabilities without corrupting its invaluable original motion priors. iMontage excels across several mainstream many-in-many-out tasks, not only maintaining strong cross-image contextual consistency but also generating scenes with extraordinary dynamics that surpass conventional scopes. Find our homepage at: https://kr1sjfu.github.io/iMontage-web/.
Authors: Yuki Kubota, Taiki Fukiage
Abstract: Deep neural networks (DNNs) are increasingly used as functional models of human vision, yet standard monocular depth estimation (MDE) benchmarks largely evaluate physical accuracy rather than behavioral alignment with humans. We introduce a cross-domain behavioral benchmark that aligns newly collected human absolute-distance judgments, physical ground truth, and model predictions at 16 locations in each of 652 KITTI and 654 NYU Depth V2 scenes. Human-model similarity was measured image-wise by partial correlation between scale-and-shift-aligned estimates while controlling for physical depth, and reference results were obtained for 69 diverse MDE models. Human error patterns were highly reliable, and most models showed positive but substantially lower similarity to humans. Human similarity, however, did not increase monotonically with metric accuracy: within each dataset, high-accuracy models converged toward similar residual error structures that remained distinct from humans, while both human-similarity scores and error-space organization were only weakly preserved across datasets. Thus, physical accuracy and human-like behavior are complementary, domain-sensitive dimensions of MDE performance. Our benchmark will provide a reusable basis for assessing future models along both dimensions.
Authors: Fr\'ed\'eric Fortier-Chouinard, Yannick Hold-Geoffroy, Valentin Deschaintre, Matheus Gadelha, Jean-Fran\c{c}ois Lalonde
Abstract: Recent progress in text-to-video generation has achieved remarkable realism, yet fine-grained control over camera motion and orientation remains elusive, especially with extreme trajectories (e.g., a 180-degree turnaround, or looking directly up or down). Existing approaches typically encode camera trajectories using relative or ambiguous representations, limiting precise geometric control and offering limited support for large rotations. We introduce GimbalDiffusion, a framework that enables camera control grounded in physical-world coordinates, using gravity as a global reference. Instead of describing motion relative to previous frames, our method defines camera trajectories in an absolute coordinate system, allowing accurate, interpretable control over camera parameters. Using panoramic 360-degree videos for training, we cover the full sphere of possible viewpoints, including combinations of extreme pitch and roll that are out-of-distribution of conventional video data. To improve camera control, we introduce null-pitch conditioning, a strategy that prevents the model from overriding camera specifications in the presence of conflicting prompt content (e.g., generating grass while the camera points toward the sky). Finally, we propose new benchmarks to evaluate gravity-aware camera-controlled video generation, assessing models' ability to generate extreme camera angles and quantify their input prompt entanglement.
Authors: Chang Cai, Hao Jiang, Xiaojun Yuan, Ying-Jun Angela Zhang
Abstract: Message-passing algorithms have been adapted for compressive imaging by incorporating various off-the-shelf image denoisers. However, these denoisers rely largely on generic or hand-crafted priors and often fall short in accurately capturing the complex statistical structure of natural images. As a result, traditional plug-and-play (PnP) methods often lead to suboptimal reconstruction, especially in highly underdetermined regimes. Recently, score-based generative models have emerged as a powerful framework for accurately characterizing sophisticated image distributions. Yet, their direct use for posterior sampling typically incurs prohibitive computational complexity. In this paper, by exploiting the close connection between score-based generative modeling and empirical Bayes denoising, we devise a message-passing framework that integrates a score-based minimum mean-squared error (MMSE) denoiser for compressive image recovery. The resulting algorithm, named score-based turbo message passing (STMP), combines the fast convergence of message passing with the expressive power of score-based generative priors. For practical systems with quantized measurements, we further propose quantized STMP (Q-STMP), which augments STMP with a component-wise MMSE dequantization module. We demonstrate that the asymptotic performance of STMP and Q-STMP can be accurately predicted by a set of state-evolution (SE) equations. Experiments on the FFHQ dataset demonstrate that STMP strikes a significantly better performance-complexity tradeoff compared with competing baselines, and that Q-STMP remains robust even under 1-bit quantization. Remarkably, both STMP and Q-STMP typically converge within 10 iterations.
Authors: Hyeonjeong Ha, Jinjin Ge, Bo Feng, Kaixin Ma, Gargi Chakraborty
Abstract: Multimodal large language models (MLLMs) have achieved impressive progress in vision-language reasoning, yet their ability to understand temporally unfolding narratives in videos remains underexplored. True narrative understanding requires grounding who is doing what, when, and where, maintaining coherent entity representations across dynamic visual and temporal contexts. We introduce NarrativeTrack, the first benchmark to evaluate narrative understanding in MLLMs through fine-grained entity-centric reasoning. Unlike existing benchmarks limited to short clips or coarse scene-level semantics, we decompose videos into constituent entities and examine their continuity via a Compositional Reasoning Progression (CRP), a structured evaluation framework that progressively increases narrative complexity across three dimensions: entity existence, entity changes, and entity ambiguity. CRP challenges models to advance from temporal persistence to contextual evolution and fine-grained perceptual reasoning. A fully automated entity-centric pipeline enables scalable extraction of temporally grounded entity representations, providing the foundation for CRP. Evaluations of state-of-the-art MLLMs reveal that models fail to robustly track entities across visual transitions and temporal dynamics, often hallucinating identity under context shifts. Open-source general-purpose MLLMs exhibit strong perceptual grounding but weak temporal coherence, while video-specific MLLMs capture temporal context yet hallucinate entities' contexts. These findings uncover a fundamental trade-off between perceptual grounding and temporal reasoning, indicating that narrative understanding emerges only from their integration. NarrativeTrack provides the first systematic framework to diagnose and advance temporally grounded narrative comprehension in MLLMs.
Authors: Souhail Hadgi, Bingchen Gong, Ramana Sundararaman, Emery Pierson, Lei Li, Peter Wonka, Maks Ovsjanikov
Abstract: Current foundation models for 3D shapes excel at global tasks (retrieval, classification) but transfer poorly to local part-level reasoning. Recent approaches leverage vision and language foundation models to directly solve dense tasks through multi-view renderings and text queries. While promising, these pipelines require expensive inference over multiple renderings, depend heavily on large language-model (LLM) prompt engineering for captions, and fail to exploit the inherent 3D geometry of shapes. We address this gap by introducing an encoder-only 3D model that produces language-aligned patch-level features directly from point clouds. Our pre-training approach builds on existing data engines that generate part-annotated 3D shapes by pairing multi-view SAM regions with VLM captioning. Using this data, we train a point cloud transformer encoder in two stages: (1) distillation of dense 2D features from visual encoders such as DINOv2 into 3D patches, and (2) alignment of these patch embeddings with part-level text embeddings through a multi-positive contrastive objective. Our 3D encoder achieves zero-shot 3D part segmentation with fast single-pass inference without any test-time multi-view rendering, while significantly outperforming previous rendering-based and feed-forward approaches across several 3D part segmentation benchmarks. Project website: https://souhail-hadgi.github.io/patchalign3dsite/
Authors: Yuteng Liu, Duanni Meng, Yimian Dai, Maoxun Yuan, Xingxing Wei, Bo li
Abstract: Infrared small target detection (IRSTD) faces significant challenges due to low signal-to-noise ratios, extremely small target sizes, and complex cluttered backgrounds. Although DETR-based detectors benefit from global context modeling, their query initialization can become unreliable in IRSTD because only a few encoder tokens correspond to targets, while the majority describe the background. We revisit this phenomenon and reveal that the target-relevant embeddings of IRST are inevitably overwhelmed by dominant background features due to the self-attention mechanism, leading to unreliable query initialization and inaccurate target localization. To address this issue, we propose SEF-DETR, a novel framework that refines query initialization through Patch-wise Spectral Screening (PSS), Frequency-Routed Examination (FRE), and Reliability-Consistency-aware Fusion (RCF). PSS encodes the radial and directional energy distributions of local Fourier spectra and aggregates overlapping patch predictions into a target-relevant density map. Guided by this map, FRE performs sparse deformable re-examination after each encoder layer, while retaining content-driven sampling to distinguish true targets from frequency-domain false alarms. RCF further re-ranks candidate queries according to spatial-frequency consistency and frequency reliability. Extensive experiments on three public IRSTD datasets demonstrate that SEF-DETR achieves superior detection performance over state-of-the-art methods with low computational overhead, providing a robust and efficient DETR-based solution for IRSTD.
Authors: Chengtao Lv, Yumeng Shi, Yushi Huang, Ruihao Gong, Shen Ren, Wenya Wang
Abstract: Advanced autoregressive (AR) video generation models have improved visual fidelity and interactivity, but the quadratic complexity of attention remains a primary bottleneck for efficient deployment. While existing sparse attention solutions have shown promise on bidirectional models, we identify that applying these solutions to AR models leads to considerable performance degradation for two reasons: isolated consideration of chunk generation and insufficient utilization of past informative context. Motivated by these observations, we propose \textsc{Light Forcing}, the \textit{first} sparse attention solution tailored for AR video generation models. It incorporates a \textit{Chunk-Aware Growth} mechanism to quantitatively estimate the contribution of each chunk, which determines their sparsity allocation. This progressive sparsity increase strategy enables the current chunk to inherit prior knowledge in earlier chunks during generation. Additionally, we introduce a \textit{Hierarchical Sparse Attention} to capture informative historical and local context in a coarse-to-fine manner. Such two-level mask selection strategy (i.e., frame and block level) can adaptively handle diverse attention patterns. Extensive experiments demonstrate that our method outperforms existing sparse attention in quality (e.g., 84.5 on VBench) and efficiency (e.g., $1.2{\sim}1.3\times$ end-to-end speedup). Combined with other efficient solutions, \textsc{Light Forcing} further achieves a $2.0{\sim}3.0\times$ end-to-end speedup across diverse GPUs (e.g., 27.4\,FPS on RTX 5090 and 33.9\,FPS on H100). Code is released via this \href{https://github.com/chengtao-lv/LightForcing}{link}.
Authors: Minggui He, Mingchen Dai, Jian Zhang, Yilun Liu, Shimin Tao, Pufan Zeng, Osamu Yoshie, Yuya Ieiri
Abstract: Vision-Language Models (VLMs) have shown promise in generating plotting code from chart images, yet achieving structural fidelity remains challenging. Existing approaches largely rely on supervised fine-tuning, encouraging surface-level token imitation rather than faithful modeling of underlying chart structure, which often leads to hallucinated or semantically inconsistent outputs. We propose Chart Specification, a structured intermediate representation that shifts training from text imitation to semantically grounded supervision. Chart Specification filters syntactic noise to construct a structurally balanced training set and supports a Spec-Align Reward that provides fine-grained, verifiable feedback on structural correctness, enabling reinforcement learning to enforce consistent plotting logic. Experiments on three public benchmarks show that our method consistently outperforms prior approaches. With only 3K training samples, we achieve strong data efficiency, surpassing leading baselines by up to 61.7% on complex benchmarks, and scaling to 4K samples establishes new state-of-the-art results across all evaluated metrics. Overall, our results demonstrate that precise structural supervision offers an efficient pathway to high-fidelity chart-to-code generation. Code and dataset are available at: https://github.com/Mighten/chart-specification-paper
Authors: Zesheng Jia, Jin Wang, Siao Liu, Lingzhi Li, Ziyao Huang, Yunjiang Xu, Jianping Wang
Abstract: Efficient domain adaptation remains a fundamental challenge for deploying multi-agent systems across diverse environments in Vehicle-to-Everything (V2X) collaborative perception. Despite the success of Parameter-Efficient Fine-Tuning (PEFT) in natural language processing and conventional vision tasks, directly applying PEFT to collaborative perception recovers only a limited portion of the performance lost to domain shift. In this work, we identify two complementary bottlenecks that limit this recovery: (i) inter-frame redundancy within a collaborative sequence, which makes the effectiveness of scarce labels sensitive to frame selection, and (ii) foreground cues that become less linearly decodable in deeper-stage representations of a frozen backbone. To address these issues, we propose FlowAdapt, a parameter-efficient framework grounded in optimal transport. Wasserstein Greedy Sampling casts frame selection as minimizing the $W_{\infty}$ distance from the sequence to the retained subset, which equals its covering radius, so a farthest first traversal returns a subset provably within twice the optimum. Progressive Knowledge Transfer then routes compressed early-stage features into the deeper stages, gating each stage-local correction by that early evidence. Extensive experiments across target domains and fusion architectures show that FlowAdapt achieves state-of-the-art adaptation performance with about 1\% trainable parameters, and maintains this lead under localization noise.
Authors: Xingyu Shao, Mengfan He, Liangzheng Sun, Chunyu Li, Ziyang Meng
Abstract: Matching downward-looking unmanned aerial vehicle (UAV) images to georeferenced satellite or aerial map tiles supports local earth observation, map-based interpretation, and coarse geo-initialization when direct positioning signals are degraded or unavailable. A major difficulty in this cross-platform matching task is the scale mismatch caused by large variations in UAV altitude. To address this problem, we propose an altitude-adaptive aerial visual place recognition framework. The method first estimates relative altitude from a single downward-looking image by transforming the input into the frequency domain and formulating altitude estimation as a regression-as-classification problem. The estimated altitude is then used to crop the query image to a canonical scale, after which a classification-then-retrieval visual place recognition module performs coarse map localization. To improve retrieval robustness under varying image quality, we further introduce a quality-adaptive margin classifier and refine the final location by weighted coordinate estimation over the top retrieved candidates. Experiments on two synthetic datasets and two real-flight datasets show that the relative altitude estimation module improves downstream retrieval performance under substantial altitude changes. With our visual place recognition module, altitude adaptation improves average R@1 and R@5 by 41.50 and 56.83 percentage points, respectively, compared with using the same retrieval pipeline without altitude normalization, and the main retrieval pipeline runs at 13.3 frames/s on the workstation used for evaluation. These results indicate that relative altitude estimation provides a useful scale prior for cross-altitude UAV-to-map image matching using georeferenced remote sensing imagery.
Authors: Zhiye Wang, Yanbo Jiang, Rui Zhou, Bo Zhang, Fang Zhang, Zhenhua Xu, Yaqin Zhang, Jianqiang Wang
Abstract: Large language models (LLMs) have shown great promise for autonomous driving. However, discretizing numbers into tokens limits precise numerical reasoning, fails to reflect the positional significance of digits in the training objective, and makes it difficult to achieve both decoding efficiency and numerical precision. These limitations affect both the processing of sensor measurements and the generation of precise control commands, creating a fundamental barrier for deploying LLM-based autonomous driving systems. In this paper, we introduce DriveCode, a novel numerical encoding method that represents numbers as dedicated embeddings rather than discrete text tokens. DriveCode employs a number projector to map numbers into the language model's hidden space, enabling seamless integration with visual and textual features in a unified multimodal sequence. Evaluated on OmniDrive, DriveGPT4, and DriveGPT4-V2 datasets, DriveCode demonstrates superior performance in trajectory prediction and control signal generation, confirming its effectiveness for LLM-based autonomous driving systems.
Authors: Zixiao Wen, Guangyao Zhou, Jiawei Li, Xiantai Xiang, Zhen Yang, Yuxin Hu, Yuhan Liu
Abstract: Satellite video object tracking (SVOT) remains fundamentally challenging due to texture scarcity, arbitrary rotation, aspect ratio changes, and severe occlusions. While recent state-of-the-art trackers excel in general scenarios, their reliance on rich appearance details or rigid spatial matching mechanisms leads to significant performance degradation in the satellite domain. To bridge this gap, we propose SiamGM, a real-time spatial-temporal unified tracking framework tailored for satellite videos. Instead of conventionally stacking modules, we synergize geometric-topological perception with temporal-kinematic prior, addressing the core limitations of SVOT. Spatially, we closely couple a Topological Attention Module (TAM) with a Geometry-Constrained Label Assignment (GCLA) method during the training phase, where a Local Graph Propagation (LGP) mechanism enforces neighborhood-consistent template-search correspondences. This joint paradigm establishes fine-grained structural correspondences and explicitly suppresses surrounding background noise. Temporally, rather than passively relying on visual cues, we propose an Online Motion Model Refinement (OMMR) strategy during the tracking phase. Gated by an online self-calibrated confidence indicator, OMMR adaptively fuse long-term stable and short-term instantaneous motion estimation for robust tracking. Evaluations on two challenging SatSOT and SV248S benchmarks confirm that SiamGM outperforms not only recent Siamese-based satellite trackers but also state-of-the-art Transformer trackers in both precision and success metrics. Notably, this highly unified architecture introduces only minor computational overhead, enabling real-time tracking at 130 frames per second (FPS).
Authors: Jeonghyeok Do, Yun Chen, Geunhyuk Youk, Munchurl Kim
Abstract: Current skeleton representation learning paradigms face distinct limitations: Contrastive Learning (CL) often overlooks fine-grained motion details, while Masked Auto-Encoders (MAE) rely on coordinate-level reconstruction. This reconstruction inherently demands dense token sequences and heavy decoders, wasting pre-training computation on discarded components and forcing downstream inference to process dense token grids. To resolve these bottlenecks, we propose SLiM (Skeleton Less is More), a compact-token framework that unifies masked feature prediction and contrastive learning via a shared encoder. By shifting the objective from raw coordinate reconstruction to decoder-free, teacher-guided feature prediction, SLiM breaks the reliance on dense tokenization and enables effective learning with a highly compact token grid. Crucially, to prevent trivial shortcut learning arising from strong inter-joint dependencies of human, we introduce Semantic Tube Masking together with Skeleton-Aware Augmentations to enforce deep skeletal-temporal reasoning and anatomical consistency. Extensive experiments across multiple downstream protocols demonstrate that SLiM achieves state-of-the-art performance while structurally reducing inference computation by 7.89$\times$ compared to dense-token MAE baselines.
Authors: Stefan Maria Ailuro (INSAIT, Sofia University "St. Kliment Ohridski"), Mario Markov (INSAIT, Sofia University "St. Kliment Ohridski"), Mohammad Mahdi (INSAIT, Sofia University "St. Kliment Ohridski"), Delyan Boychev (INSAIT, Sofia University "St. Kliment Ohridski"), Luc Van Gool (INSAIT, Sofia University "St. Kliment Ohridski"), Danda Pani Paudel (INSAIT, Sofia University "St. Kliment Ohridski")
Abstract: Vision-Language Models (VLMs) adapted to remote sensing rely heavily on domain-specific image-text supervision, yet high-quality annotations for satellite and aerial imagery remain scarce and expensive to produce. Prevailing pseudo-labeling pipelines address this gap by distilling knowledge from large frontier models, but this dependence on large teachers is costly, limits scalability, and caps achievable performance at the ceiling of the teacher. We propose OSMDA: a self-contained domain adaptation framework that eliminates this dependency. Our key insight is that a capable base VLM can serve as its own annotation engine: by pairing aerial images with rendered OpenStreetMap (OSM) tiles, we leverage optical character recognition and chart comprehension capabilities of the model to generate captions enriched by OSM's vast auxiliary metadata. The model is then fine-tuned on the resulting corpus with satellite imagery alone, yielding OSMDA-VLM, a domain-adapted VLM that requires no manual labeling and no stronger external VLM teacher. We conduct exhaustive evaluations spanning six zero-shot and five in-distribution benchmarks across vision-language tasks, where OSMDA leads to substantial improvement. We further compare against nine competitive baselines, demonstrating that our method achieves superior overall performance, while being substantially cheaper to train than teacher-dependent alternatives. These results suggest that, given a strong foundation model, alignment with crowd-sourced geographic data is a practical and scalable path towards remote sensing domain adaptation. Dataset and model weights will be made publicly available upon acceptance.
Authors: Ye Pan, Chi Kit Wong, Yuanhuiyi Lyu, Hanqian Li, Chenfei Liao, Jiahao Huo, Lutao Jiang, Zixin Zhang, Jiacheng Chen, Yuqian Fu, Xu Zheng
Abstract: Egocentric video provides a natural modality for studying human behavior, but conventional visual understanding captures mainly observable scenes, objects, and actions rather than the latent goals that organize them. Existing intent benchmarks typically focus on coarse event-level goals and overlook how intent evolves across procedural steps. We introduce EgoIntent, a step-level intent-understanding benchmark comprising 3,014 pre-outcome micro-steps from 32 egocentric videos across 15 indoor and outdoor daily-life scenarios. Each step is manually annotated along three complementary dimensions: Local Intent (What), the immediate goal; Procedural Intent (Why), the role of the step in the broader procedure; and Next-Plan (Next), the action most likely to follow. Multiple rounds of human review refine temporal boundaries and annotation quality. We evaluate 15 multimodal large language models using reference-based scores and complementary reference-free diagnostics. Controlled studies on four representative models show that only one model gains significantly from correct temporal order, while a single boundary frame outperforms the full ordered clip for three models. Step-only input performs best for all four models, and adding 15 seconds of history significantly degrades three. Revealing the current outcome improves Local Intent by 7.81 points, while revealing the following step improves Next-Plan by 13.17 points. These findings indicate that current models can achieve strong intent-prediction scores through static boundary cues without robustly exploiting temporal order or procedural history.
Authors: Mattia Bernardi, Chiara Cappellino, Matteo Mosconi, Enver Sangineto, Angelo Porrello, Simone Calderara
Abstract: Although recent Open-Vocabulary Object Detection architectures, such as Grounding DINO, demonstrate strong zero-shot capabilities, their performance degrades significantly under domain shifts. Moreover, many domains of practical interest, such as nighttime or foggy scenes, lack large annotated datasets, preventing direct fine-tuning. In this paper, we introduce Aligned Basis Relocation for Adaptation(ABRA), a method that transfers class-specific detection knowledge from a labeled source domain to a target domain where no training images containing these classes are accessible. ABRA formulates this adaptation as a geometric transport problem in the weight space of a pretrained detector, aligning source and target domain experts to transport class-specific knowledge. Extensive experiments across challenging domain shifts demonstrate that ABRA successfully teleports class-level specialization under multiple adverse conditions. Our code will be made public upon acceptance.
Authors: Jinguang Tong, Jinbo Wu, Kaisiyuan Wang, Zhelun Shen, Xuan Huang, Mochu Xiang, Xuesong Li, Yingying Li, Haocheng Feng, Chen Zhao, Hang Zhou, Wei He, Chuong Nguyen, Jingdong Wang, Hongdong Li
Abstract: Human-Object Interaction (HOI) video reenactment aims to transfer the interaction dynamics of a source video to a novel target object while preserving realistic hand-object coordination. Existing methods typically rely on sparse 2D motion controls and monocular references, which are insufficient for complex out-of-plane motion and large viewpoint changes. We present MVHOI, a two-stage framework combining implicit motion extraction, 3D-aware multi-view reasoning, and video generation. In the first stage, a motion extractor encodes object dynamics into implicit motion descriptors. Conditioned on these descriptors, our Motion-Driven Object Prior (MDOP) module queries a 3D foundation model over multi-view references of the target object and autoregressively predicts coarse object anchors, a sequence of images that track the object's evolving orientation and appearance under the source motion without any explicit pose estimation. In the second stage, a DiT-based video generation model uses these anchors as structural guidance and the multi-view references as appearance guidance. We further reuse cross-view attention from MDOP as a soft attention bias to reduce reference-view confusion. For long videos, a cross-iterative inference strategy refreshes subsequent object priors using refined video outputs. Experiments demonstrate consistent improvements over state-of-the-art methods in object fidelity, motion consistency, visual quality, and interaction realism.
Authors: Hanze Jia, Chunshi Wang, Yuxiao Yang, Zhonghua Jiang, Yawei Luo, Shuainan Ye, Tan Tang
Abstract: 3D fragment reassembly estimates the rigid pose of each fragment to recover a complete object from unordered point clouds or meshes. The task becomes increasingly challenging as the fragment count grows, since irregular fragments provide weak semantic cues and admit rapidly increasing numbers of plausible contact relations and global configurations. We propose Structure-Aware Reassembly (SARe), a generative framework that integrates query-aligned local geometry and task-native structural supervision into point-flow assembly. SARe-Gen conditions each transported surface query on a local latent and jointly supervises intermediate flow tokens with query-level fracture-region and fragment-level contact targets. Because these heads are optimized together with flow matching, structural supervision directly shapes the representations that drive coordinate transport, without requiring additional reassembly-specific pretraining or a separate teacher-alignment stage. At inference time, SARe-Refine geometrically verifies predicted relations and uses reliable local subassemblies to guide a second sampling pass, reinforcing consistent regions while resampling uncertain fragments. We evaluate SARe across three settings, including synthetic fractures, simulated fractures from scanned real objects, and scans of physically fractured objects. The results demonstrate state-of-the-art performance, with higher part accuracy and more graceful degradation in challenging many-fragment settings.
Authors: Yuqin Lu, Haofeng Liu, Yang Zhou, Yihua Dai, Guiqing Li, Shengfeng He, Jun Liang
Abstract: Diffusion models provide powerful priors for 2D image completion, but these priors are learned on bounded planar images and do not transfer directly to $360^\circ$ panoramas. Perspective observations and spherical panoramas differ in both projective geometry and topology: viewpoint-dependent distortion complicates spatial correspondence, while Equirectangular Projection (ERP) panoramas exhibit intrinsic $S^1$ periodicity that standard Euclidean architectures do not preserve. We present Gimbal360, a unified framework that adapts planar diffusion priors to spherical panoramic completion by standardizing these geometric and topological structures. Our Canonical Viewing Space expresses projective distortion as a fixed function of latitude, providing a consistent interface between perspective inputs and spherical panoramas. To map unposed in-the-wild images into this space, Differentiable Projective Canonicalization projects a dense correspondence field onto a 3-DoF rigid projection manifold without requiring camera parameters at inference. We further introduce Topologically Equivariant Generation, which enforces latent shift equivariance to preserve continuity across the periodic ERP boundary. Together, these designs allow diffusion to operate in a representation whose geometry and topology are explicitly aligned with the spherical domain. We also introduce Horizon360, a curated large-scale dataset of gravity-aligned panoramic environments. Extensive experiments show that Gimbal360 achieves state-of-the-art visual fidelity and seam continuity in $360^\circ$ scene completion.
Authors: Minh-Quan Viet Bui, Jaeho Moon, Munchurl Kim
Abstract: While 3D Vision Foundation Models (3DVFMs) have demonstrated remarkable zero-shot capabilities in visual geometry estimation, their direct application to generalizable novel view synthesis (NVS) remains challenging. In this paper, we propose AirSplat, a novel training framework that effectively adapts the robust geometric priors of 3DVFMs into high-fidelity, pose-free NVS. Our approach introduces two key technical contributions: (1) Self-Consistent Pose Alignment (SCPA), a training-time feedback loop that ensures pixel-aligned supervision to resolve pose-geometry discrepancy; and (2) Rating-based Opacity Matching (ROM), which leverages the local 3D geometry consistency knowledge from a sparse-view NVS teacher model to filter out degraded primitives. Experimental results on large-scale benchmarks demonstrate that our method significantly outperforms state-of-the-art pose-free NVS approaches in reconstruction quality. Our AirSplat highlights the potential of adapting 3DVFMs to enable simultaneous visual geometry estimation and high-quality view synthesis.
Authors: Daojie Peng, Fulong Ma, Jun Ma
Abstract: Vision-Language Navigation (VLN) requires an embodied agent to navigate complex environments by following natural language instructions, which typically demands tight fusion of visual and language modalities. Existing VLN methods often convert raw images into visual tokens or implicit features, requiring large-scale visual pre-training and suffering from poor generalization under environmental variations (e.g., lighting, texture). To address these issues, we propose SOL-Nav (Structured Observation Language for Navigation), a novel framework that translates egocentric visual observations into compact structured language descriptions for efficient and generalizable navigation. Specifically, we divide RGB-D images into a NxN grid, extract representative semantic, color, and depth information for each grid cell to form structured text, and concatenate this with the language instruction as pure language input to a pre-trained language model (PLM). Experimental results on standard VLN benchmarks (R2R, RxR) and real-world deployments demonstrate that SOL-Nav significantly reduces the model size and training data dependency, fully leverages the reasoning and representation capabilities of PLMs, and achieves strong generalization to unseen environments.
Authors: Chengyin Hu, Jiaju Han, Xuemeng Sun, Qike Zhang, Luwei Yang, Lehan Sun, Jiahuan Long, Yiwei Wei, Jiujiang Guo
Abstract: Vision-language models (VLMs) share visual-textual representations across zero-shot classification, image captioning, and visual question answering (VQA), creating a pathway through which subtle perturbations may cause failures across tasks. We introduce X-shaped Sparse Pixel Attack (XSPA), a structured attack that restricts perturbations to two intersecting diagonal lines. Within this fixed support, XSPA jointly optimizes a classification objective, cross-task semantic guidance, perturbation magnitude, and linewise smoothness. Under the default setting, it changes about 1.04\% of image pixels. On COCO, XSPA reduces zero-shot accuracy by 52.33 points on OpenAI CLIP ViT-L/14 and 67.00 points on OpenCLIP ViT-B/16; GPT-4-based caption consistency and VQA correctness decrease by up to 58.60 and 44.25 points, respectively. Matched-budget experiments show that an adaptive saliency Top-k support achieves higher attack success than the fixed X-shaped support, while XSPA provides better SSIM and LPIPS but worse PSNR and MAE. Same-image analysis finds no three-task targeted success, indicating that strong surrogate classification effects do not reliably translate into targeted output-space failures. These results position XSPA as a controlled stress test for studying how sparse fixed geometry affects VLM robustness, rather than as a universally superior attack.
Authors: Eugene Lee, Ting-Yu Chang, Jui-Huang Tsai, Jiajie Diao, Chen-Yi Lee
Abstract: The field of computer vision has experienced significant advancements through scalable vision encoders and multimodal pre-training frameworks. However, existing approaches often treat vision encoders and large language models (LLMs) as independent modules, limiting the integration of hierarchical visual features. In this work, we propose HIVE (Hierarchical Pre-Training of Vision Encoders), a novel framework that enhances vision-language alignment by introducing hierarchical cross-attention between the vision encoder and LLM. Unlike conventional methods that flatten image embeddings, HIVE enables structured feature fusion across multiple layers, improving gradient flow and representation learning. To optimize this interaction, we introduce a three-stage training strategy that progressively aligns the vision encoder with the LLM, ensuring stable optimization and effective multimodal fusion. Empirical evaluations demonstrate that HIVE achieves superior performance not only in image classification but also on various vision-language tasks, outperforming self-attention-based methods in benchmarks such as MME, GQA, OK-VQA, and ScienceQA. Our results highlight the benefits of hierarchical feature integration, paving the way for more efficient and expressive vision-language models.
Authors: Li He, Longtai Zhang, Wenqiang Zhang, Yan Wang, Lizhe Qi
Abstract: Text-to-image diffusion models achieve high visual fidelity, yet fine-grained affective control remains difficult because textual emotion cues often fail to specify the visual perceptual factors underlying affective expression. Existing visual-affect datasets are likewise often limited to discrete labels, specific domains, or limited supervision of perceptual attributes. We introduce EmoScene, a large-scale dual-space dataset for controllable affective image generation, containing 1.2M images across more than 300 scene categories. Its affective space jointly represents discrete emotions and continuous valence--arousal--dominance (VAD), its perceptual space records measurable appearance attributes, and contextual descriptions ground both in scene semantics. EmoScene combines multi-model annotation with human-in-the-loop quality control. A random audit of 30,519 images yields 91.22\% agreement on discrete emotion labels, while an independent multi-rater evaluation yields Fleiss' $\kappa=0.85$. After controlling for source and scene composition, affective dimensions and perceptual attributes exhibit stable associations across data sources, reflecting statistical tendencies rather than deterministic visual rules. To demonstrate the dataset's utility, we further develop AffectCtrl, which learns residuals in the conditioning space of frozen diffusion models to support categorical emotion generation and continuous control over VAD, brightness, and saturation. AffectCtrl achieves 85.75\% categorical emotion accuracy, outperforms EmotiCrafter in valence and arousal control under a shared evaluation protocol, and obtains Pearson correlations of 0.673--0.765 across all five continuous axes. These results demonstrate that EmoScene provides a scalable data foundation for analyzing and controlling affective expression in visual generation.
Authors: Shubo Lin, Xuanyang Zhang, Wei Cheng, Weiming Hu, Gang Yu, Jin Gao
Abstract: Despite advancements in generating visually stunning content, video diffusion models (VDMs) often yield physically inconsistent results due to pixel-only reconstruction. To address this, we propose MMPhysVideo, the first study to enhance physical plausibility in video generation through joint multimodal modeling. We recast perceptual cues, specifically semantics, geometry, and spatio-temporal trajectories, into a unified pseudo-RGB format, enabling VDMs to directly capture complex physical dynamics. To mitigate cross-modal interference, we propose a Bidirectionally Controlled Teacher architecture, which utilizes parallel branches to fully decouple RGB and perception processing and adopts two zero-initialized control links to gradually establish pixel-wise consistency. For inference efficiency, the teacher's physical prior is distilled into a single-stream student model via representation alignment. Furthermore, we present MMPhysPipe, an end-to-end data curation and annotation pipeline tailored for constructing physics-rich multimodal datasets. MMPhysPipe employs a vision-language model (VLM) guided by a chain-of-visual-evidence rule to pinpoint physical subjects, enabling expert models to extract multi-granular perceptual information. Without additional inference costs, MMPhysVideo consistently improves physical plausibility of advanced models on the Videophy and PhyGenbench benchmarks and achieves superior performance among existing methods.
Authors: Rixiang Ni, Boyang Li, Jun Chen, Yonghao Li, Feiyu Ren, Yuji Wang, Haoyang Yuan, Wujiao He, Wei An
Abstract: Infrared small target detection (IRSTD) aims to separate small targets from clutter backgrounds. Extensive research is dedicated to the pixel-level supervision-guided "encoder-decoder" segmentation paradigm. Although having achieved promising performance, they neglect the fact that small targets only occupy a few pixels and are usually accompanied with blurred boundary caused by clutter backgrounds. Based on this observation, we argue that the first principle of IRSTD should be target localization instead of separating all target region accompanied with indistinguishable background noise. In this paper, we reformulate IRSTD as a centroid regression task and propose a novel Single-Point Supervision guided Infrared Probabilistic Response Encoding method (namely, SPIRE), which is indeed challenging due to the mismatch between reduced supervision network and equivalent output. Specifically, we first design a Point-Response Prior Supervision (PRPS), which transforms single-point annotations into probabilistic response map consistent with infrared point-target response characteristics, with a High-Resolution Probabilistic Encoder (HRPE) that enables encoder-only, end-to-end regression without decoder reconstruction. By preserving high-resolution features and increasing effective supervision density, SPIRE alleviates optimization instability under sparse target distributions. Finally, extensive experiments on various IRSTD benchmarks, including SIRST-UAVB and SIRST4 demonstrate that SPIRE achieves competitive target-level detection performance with consistently low false alarm rate (Fa) and significantly reduced computational cost. Code is publicly available at: https://github.com/NIRIXIANG/SPIRE-IRSTD.
Authors: Yushuo Zhang, Yu Cheng, Yongkang Hu, Jiuan Zhou, Jiawei Chen, Zhaoxia Yin
Abstract: Facial forgery techniques are advancing rapidly, posing severe threats to public trust and information security while imposing higher demands on the continual adaptation of DeepFake detection models. Although continual learning enables models to adapt to emerging forgery methods, existing approaches still face two key bottlenecks. On the one hand, they lack sufficient feature representation capacity to handle increasingly diverse and complex forgery traces. On the other hand, continual adaptation to new forgery distributions leads to severe catastrophic forgetting of prior knowledge, which substantially degrades detection performance. To address these issues, we propose Face-D(^2)CL, a framework for facial DeepFake detection. It leverages multi-domain synergistic representation to fuse spatial and frequency-domain features, enabling comprehensive capture of diverse forgery traces. Additionally, it employs a dual continual learning mechanism that combines Real/Fake-aware Elastic Weight Consolidation (RF-EWC) and Domain-wise Orthogonal Gradient Constraint (D-OGC). RF-EWC distinguishes the parameter importance for real versus fake samples, while D-OGC ensures that updates to task-specific expert modules do not interfere with previously learned knowledge. This synergy allows the model to achieve a dynamic balance between robust anti-forgetting capabilities and agile adaptability to emerging facial forgery paradigms, all without relying on historical data replay. Extensive experiments demonstrate that our method surpasses current state-of-the-art (SOTA) approaches in both stability and plasticity, achieving a 60.7% relative reduction in the average detection error rate. On unseen forgery domains, it further improves the average detection AUC by 7.9% compared to the current SOTA method.
Authors: Ruizhi Zhang, Ye Huang, Yuangang Pan, Chuanfu Shen, Zhilin Liu, Ting Xie, Haijun Lei, Lixin Duan
Abstract: While artificial intelligence has mastered structured games like chess and Go, vision-language agents still struggle in visually-driven 3D games without access to game states. Existing game environments typically evaluate a fixed agent configuration, rather than an agent's ability to improve its configuration across consecutive episodes of the same task---a paradigm known as test-time learning (TTL). Furthermore, current TTL methods typically optimize single modalities---such as text prompts or actions---in isolation, ignoring the synergy between perception, reasoning, and control. To bridge these gaps, we first introduce \textbf{PokeGym}, a long-horizon benchmark built upon the 3D open-world game Pok\'emon Legends: Z-A, where agents act from visual observations without access to game states, designed to evaluate an agent's ability to learn and adapt across consecutive episodes of the task. To tackle this challenging environment, we propose Graph-Guided Evolutionary Multimodal Agent Configuration (\textbf{G-EvoMAC}), a graph-guided framework that jointly optimizes visual perception, strategy, and action set synergistically. Extensive experiments show that G-EvoMAC achieves a 60.18\% average success rate on PokeGym, outperforming the strongest baseline by over 11 percentage points, validating the power of cross-modal co-evolution.
Authors: Zheng Jiang, Yiming Chen, Nan He, Jiahui Chen, Chaoyang Li, Houde Qian, Lifeng Sun
Abstract: Recent multimodal large language models (MLLMs) support Thinking with Images, invoking visual tools such as zooming and cropping to inspect image regions during inference. Yet these systems remain brittle in fine-grained reasoning: to acquire a decisive detail, a model must ground its attention on the correct region, but knowing which region is correct presupposes having already observed that detail. We identify this circular dependency as the grounding paradox, show that grounding errors are rarely self-corrected within a single trajectory---once a misleading region is inspected, all subsequent reasoning conditions on that observation and the error propagates to the final answer---and observe that because each trajectory constructs its own evidence, answer-level aggregation discards the very information that distinguishes trajectories. We propose Test-Time Scaling over Perception (TTSP), a closed-loop framework that treats perception as the unit of scalable inference and allocates compute along two axes: Entropy-Gated Perceptual Exploration samples diverse trajectories and uses critical-token entropy to withhold evidence the model cannot commit to, while Evidence-Guided Iterative Refinement distills validated observations into a correctable Evidence Ledger that steers later rounds to re-inspect unresolved regions. Across high-resolution and general multimodal benchmarks, TTSP consistently outperforms strong test-time scaling baselines, while improving grounding quality with favorable token efficiency.
Authors: Huanzhen Wang, Ziheng Zhou, Jiaqi Song, Li He, Yunshi Lan, Yan Wang, Wenqiang Zhang
Abstract: Dynamic facial expression recognition in the wild remains challenging due to data scarcity and long-tail distributions, which hinder models from effectively learning the temporal dynamics of scarce emotions. To address these limitations, we propose ARGen, an Affect-Reinforced Generative Augmentation Framework that enables data-adaptive dynamic expression generation for robust emotion perception. ARGen operates in two stages: Affective Semantic Injection (ASI) and Adaptive Reinforcement Diffusion (ARD). The ASI stage establishes affective knowledge alignment through facial Action Units and employs a retrieval-augmented prompt generation strategy to synthesize consistent and fine-grained affective descriptions via large-scale visual-language models, thereby injecting interpretable emotional priors into the generation process. The ARD stage integrates text-conditioned image-to-video diffusion with reinforcement learning, introducing inter-frame conditional guidance and a multi-objective reward function to jointly optimize expression naturalness, facial integrity, and generative efficiency. Extensive experiments on both generation and recognition tasks verify that ARGen substantially enhances synthesis fidelity and improves recognition performance, establishing an interpretable and generalizable generative augmentation paradigm for vision-based affective computing.
Authors: Yiwei Zhao, Yi Zheng, Huapeng Su, Jieyu Lin, Stefano Ambrogio, Cijo Jose, Michael Ramamonjisoa, Patrick Labatut, Barbara De Salvo, Chiao Liu, Phillip B. Gibbons, Ziyun Li
Abstract: Always-on contextual AI runs language-aligned vision foundation models (VFMs) on edge devices, where the on-device model is the dominant continuous compute cost under strict latency and power limits. Due to an observed low-frequency shift in scene context and its relevant vocabulary, we present AdaDINO, an adaptive framework that makes on-device VFM inference efficient by matching execution to the current scene and task. We build on a known phenomenon, that the accuracy drop of shrinking model sizes depends on the task, and turn it into task-level adaptive execution. AdaDINO integrates neural architecture search (NAS) into a language-aligned VFM backbone distilled from DINOv2, training a single family of subnets for efficient execution during runtime. A multimodal large language model (LLM) on the cloud, invoked at low frequency, refines the candidate class set from scene context, while a learned selector activates the least-cost subnet predicted to retain a target fraction of accuracy. With the backbone and semantic pipeline held fixed, learned selection alone reduces average compute by $37\%$ over the best fixed subnet at equal segmentation accuracy. Across zero-shot classification and open-vocabulary segmentation, AdaDINO establishes a strong accuracy-efficiency frontier, improving over evaluated models of comparable sizes by up to $7.9\%$ in acc@1 on IN1K and $5.2\%$ mIoU on ADE20K, and reducing average FLOPs by up to $74.9\%$ at similar accuracy.
Authors: Afshan Hashmi
Abstract: Diabetic retinopathy (DR) is a leading cause of preventable blindness, and automated grading could extend screening capacity. However, most reported DR models are validated only on the dataset they were trained on, leaving their behaviour under real screening variability unmeasured. This study presents a dual-resolution grading framework and quantifies how far performance falls when the imaging domain shifts. Two EfficientNet backbones process complementary views of each fundus image: B0 receives Ben Graham-normalised input at 224x224, emphasising vascular structure, while B3 receives CLAHE-enhanced input at 300x300, emphasising focal lesions. A learnable attention gate fuses the branches per image, and an ordinal binary-decomposition head models severity as an ordered scale rather than as unordered categories. Training used a combined set of 4,149 images (APTOS 2019, n = 2,929; Messidor-2 training portion, n = 1,220); evaluation used a held-out APTOS split (n = 733) and a Messidor-2 test set (n = 524) excluded from training and from all model selection. Quadratic weighted kappa was 0.882 (95% CI 0.853-0.906) on APTOS and 0.679 (95% CI 0.613-0.735) on Messidor-2 for this run, a significant gap of 0.202 (95% CI 0.142-0.273); across three random seeds the held-out kappa was 0.689 +/- 0.021. Critically, accuracy fell 19.3 points while 93.7% of predictions stayed within one grade of reference: ordering survives domain shift, threshold placement does not. Referable-DR sensitivity fell from 0.879 to 0.620.
Authors: Wei Yao, Haohan Ma, Hongwen Zhang, Liangjun Xing, Zhile Yang, Yuanjun Guo, Yunlian Sun, Yebin Liu
Abstract: Controllable cooperative humanoid manipulation is a fundamental yet challenging problem for embodied intelligence, due to severe data scarcity, complexities in multi-agent coordination, and limited generalization across objects. In this paper, we present SynAgent, a unified framework that enables scalable and physically plausible cooperative manipulation by leveraging Solo-to-Cooperative Agent Synergy to transfer skills from single-agent human-object interaction to multi-agent human-object-human scenarios. To maintain semantic integrity during motion transfer, we introduce an interaction-preserving retargeting method based on an Interact Mesh constructed via Delaunay tetrahedralization, which faithfully maintains spatial relationships among humans and objects. Building upon this refined data, we propose a single-agent pretraining and adaptation paradigm that distills synergistic collaborative behaviors from abundant single-human data through decentralized training and multi-agent PPO. Finally, we develop a trajectory-conditioned generative policy using a conditional VAE, trained via multi-teacher distillation from motion imitation priors to achieve stable and controllable object-level trajectory execution. Extensive experiments demonstrate that SynAgent significantly outperforms existing baselines in both cooperative imitation and trajectory-conditioned control, while generalizing across diverse object geometries. Codes and data will be available after publication. Project Page: https://yw0208.github.io/synagent/
Authors: Haokun Wen, Xuemeng Song, Haoyu Zhang, Weili Guan, Xiangyu Zhao, Liqiang Nie
Abstract: Composed image retrieval, multi-turn composed image retrieval, and composed video retrieval all share a common paradigm: composing the reference visual with modification text to retrieve the desired target. Despite this shared structure, the three tasks have been studied in isolation, with no prior work proposing a unified framework, let alone a zero-shot solution. In this paper, we propose UniCVR, the first unified zero-shot composed visual retrieval framework that jointly addresses all three tasks without any task-specific human-annotated data. UniCVR strategically combines two complementary strengths: Multimodal Large Language Models (MLLMs) for compositional query understanding and Vision-Language Pre-trained (VLP) models for structured visual retrieval. Concretely, UniCVR operates in two stages. In Stage~I, we train the MLLM as a compositional query embedder via contrastive learning on a curated multi-source dataset of approximately 3.5M samples, bridging the heterogeneous embedding spaces between the MLLM and the frozen VLP gallery encoder. A cluster-based hard negative sampling strategy is proposed to strengthen contrastive supervision. In Stage~II, we introduce an MLLM-guided dual-level reranking mechanism that applies adaptive budgeted subset scoring to a small number of top-ranked candidates, and then exploits the resulting relevance signals through a dual-level re-scoring scheme, producing more accurate final rankings with minimal computational overhead. Extensive experiments across five benchmarks covering all three tasks demonstrate that UniCVR achieves cutting-edge performance, validating its effectiveness and generalizability.
Authors: Ying Zang, Xuanyi Liu, Yidong Han, Deyi Ji, Chaotao Ding, Yuanqi Hu, Qi Zhu, Xuanfu Li, Jin Ma, Lingyun Sun, Tianrun Chen, Lanyun Zhu
Abstract: Reconstructing dynamic 4D scenes from monocular videos is a fundamental yet challenging task. While recent 3D foundation models provide strong geometric priors, their performance significantly degrades in dynamic environments. This degradation stems from a fundamental tension: the inherent coupling of camera ego-motion and object motion within global attention mechanisms. In this paper, we propose a novel, training-free progressive decoupling framework that disentangles dynamics from statics in a principled, coarse-to-fine manner. Our core insight is to resolve the tension by first stabilizing the camera pose, followed by geometric refinement. Specifically, our approach consists of three synergistic components: (1) a Dynamic-Mask-Guided Pose Decoupling module that isolates pose estimation from dynamic interference, yielding a stable motion-free reference frame; (2) a Topological Subspace Surgery mechanism that orthogonally decomposes the depth manifold, safely preserving dynamic objects while injecting refined, mask-aware geometry into static regions; and (3) an Information-Theoretic Confidence-Aware Fusion strategy that formulates depth integration as a heteroscedastic Bayesian inference problem, adaptively blending multi-pass predictions via inverse-variance weighting. Extensive experiments on standard 4D reconstruction benchmarks demonstrate that our method achieves consistent and substantial improvements across principal point-cloud metrics. Notably, our approach shows competitive performance in robust 4D scene reconstruction without requiring fine-tuning, suggesting the potential of mathematically grounded dynamic-static disentanglement.
Authors: Yuchen Li, Ziru Wei, Zhen Zhao, Yi Liu, Luping Zhou
Abstract: Medical referring image segmentation (MRIS) predicts lesion masks from medical images and natural-language referring expressions, but acquiring paired pixel-level annotations and referring texts is costly. Semi-supervised learning (SSL) can alleviate this burden by exploiting unlabeled data, yet its effectiveness depends on preserving image--text alignment under strong perturbations. Existing SSL methods for referring segmentation rely on independent or simple multimodal perturbations (e.g., left--right flips), while stronger augmentations such as CutMix remain largely unexplored because they can disrupt cross-modal correspondence. We propose Semi-MedRef, a teacher--student SSL framework that explicitly preserves alignment between medical images and positional language through three complementary components: T-PatchMix, an alignment-preserving cross-modal augmentation that synchronizes patch mixing with positional-language and pseudo-mask updates; PosAug, a position-aware text augmentation that regularizes reliance on positional expressions; and Positional Affinity Contrastive Learning (PACL), which exploits coarse positional cues to construct region-aware supervision through anatomically weighted soft positives, encouraging anatomically grounded cross-modal representation learning. Experiments on QaTa-COV19 and MosMedData+ demonstrate that Semi-MedRef consistently outperforms state-of-the-art fully supervised and semi-supervised MRIS methods across all label regimes.
Authors: Ziyan Huang, Lang Wu, Hongji Wang, Yifei Liu, Dongliang Tang, Hongqiao Wang
Abstract: Low-light, long-exposure defocus deblurring remains a challenging problem due to the simultaneous presence of severe blur and complex biased noise. Existing methods typically rely on simplified noise assumptions, which limits their effectiveness under realistic imaging conditions. In this work, we propose Physen-Noise2Noise, a self-supervised deblurring framework guided by the physical model of defocus imaging, which leverages noisy multi-frame observations without requiring clean reference images. Unlike conventional Noise2Noise-based approaches that assume zero-mean noise, we derive a frequency-domain constraint inherent to the defocus imaging process and incorporate it into the learning framework via a learnable noise bias parameter. In addition, a multi-frame noisy initialization strategy is introduced to suppress complex biased noise prior to deblurring, providing a more stable starting point for reconstruction. This formulation explicitly models biased noise and enables joint bias correction and high-frequency detail recovery during training. Furthermore, we develop a pretrain-finetune variant to enhance robustness and generalization under challenging noise conditions. Extensive experiments on both simulation and real-world datasets demonstrate that the proposed method consistently outperforms state-of-the-art self-supervised approaches for defocus deblurring in the presence of complex biased noise.
Authors: Sangeeta ., Maddikuntla Sai Prajwal, Debi Prosad Dogra, Kamalakar Vijay Thakare, Hyungjoo Jung, Ig-Jae Kim, Heeseung Choi
Abstract: Women's safety and security are paramount for a modern society. Often, crimes scenes get recorded through low-resolution CCTV cameras limiting the efficiency of video anomaly detection (VAD) models. Despite substantial progress in VAD research, women-centric anomalies are still underrepresented in datasets as well as in models. Existing datasets primarily cover well-lit, high-resolution and close-shot videos that are inadequate to tackle critical anomalies such as chain snatching, stalking, inappropriate touch, and other subtle forms of crime against women. To address this, we present a new benchmark, referred to as ExtrAnom. It contains 1001 videos (both anomalies and normal) with four textual annotations; one human-generated and three LLM-generated. The videos are arranged in 5 different categories of crimes. The dataset comprises low-light (8%), low-resolution (13%), long-shot (15%), and daytime (64%) anomaly videos. It includes stalking (3.9%), chain snatching (17.6%), kidnapping (7.3%), assassinations (2.3%), harassment (18.9%), and normal (50%) videos. It is possible to perform cross-modal and VLM-based validations using ExtrAnom. We have benchmarked it against popular unimodal and multi-modal VAD datasets (e.g., XD-Violence, UCF-Crime, and UCA) and SOTA methods. Experiments reveal that existing datasets are insufficient to deal with women-centric anomalies. We believe ExtrAnom can fill this critical gap in VAD research.
Authors: Yihang Tao, Yu Guo, Senkang Hu, Yanan Ma, Zihan Fang, Sam Kwong, Yuguang Fang
Abstract: Connected and autonomous driving (CAD) systems leverage vehicle-to-vehicle (V2V) communication for multi-agent collaborative perception, yet remain constrained by scarce annotated real-world V2V datasets and limited generalization across diverse driving conditions. While image generation offers a feasible solution for data augmentation, existing single-vehicle multi-view generation frameworks face two key challenges in multi-agent settings: (1) the expanded learning objective degrades generation quality, and (2) dynamic inter-agent variation hinders consistency modeling for physical attributes (e.g., color, category) of jointly observed objects. To bridge this gap, we propose V2VCrafter, the first framework for generating controllable and realistic multi-view driving images across vehicles. For effective learning, we develop a progressive multi-agent diffusion model based on a single-agent backbone, using neighboring agents' latent states to progressively guide single-to-multi-agent generation. To address cross-vehicle inconsistency, we further propose a cross-agent attention module that leverages a collaboration view graph and learnable jointly observed object representations to model dynamic cross-vehicle camera view relationships. Experiments on real-world V2X-Real dataset show that V2VCrafter generates high-fidelity, controllable, and consistent street views across vehicles, thereby effectively enhancing downstream collaborative 3D object detection tasks.
Authors: Xin Dong, Weijian Deng, Lihan Zhang, Tianru Dai, Wenfeng Deng, Yansong Tang
Abstract: Reconstructing interaction-ready 3D worlds is essential for physical simulation, virtual reality, robotics, and autonomous driving. However, existing methods mainly optimize static and holistic visual fidelity, with limited support for multi-object interaction. We argue that an interaction-ready reconstruction should anticipate potential scene changes and preserve geometric completeness, visual quality, multi-object spatial relationship, and physical plausibility under potential interactions. To this end, motivated by the causal intervention, we propose CA-World, an efficient framework that integrates counterfactual alignment learning into a decoupling-reintegration reconstruction pipeline. Specifically, we formulate foreground-background decoupling as a visual intervention, separate object generation and background inpainting as counterfactual generation, and scene reintegration as an inverse intervention. According to counterfactual consistency, reversing the intervention should recover the factual world, motivating three alignment objectives between the reintegrated and original scenes: appearance, spatial, and physical consistency. This formulates interaction-ready reconstruction as counterfactual alignment learning with direct supervision. Moreover, leveraging the locality of object-level interventions, CA-World constrains counterfactual states using the observed scene, enabling efficient and coherent reintegration without jointly optimizing all object states, thereby reducing computational cost and error accumulation. Experiments on object completeness, spatial accuracy, outdoor background completion, rendering quality, simulated dynamics, and downstream applications demonstrate the effectiveness of CA-World. Project page: https://chnxindong.github.io/ca-world/.
Authors: Yefeng Wu
Abstract: RGB-infrared detectors typically discard the statistics generated during cross-modal fusion, leaving downstream modules unaware of whether the current interaction is reliable. We propose to extract a parameter-free, 7-dimensional spectral reliability descriptor -- summarizing band energy, amplitude ratio, phase consistency, and cross-modal correlation -- and to reuse it beyond the fusion stage. The descriptor drives both Spectral Reliability Fusion (SRF), which gates a spectral residual against a conservative spatial base, and Reliability-Conditioned Expert Routing (RCER), which combines the descriptor with pooled content to steer sparse post-fusion experts. Under matched ablations, descriptor-aware gating improves mAP50 over content-only adaptive gating; a $2{\times}2$ factorial analysis further shows that descriptor-conditioned routing provides the larger marginal gain over expert architecture alone at near-equal parameter count. Under six synthetic degradations on DroneVehicle, average retention rises to 95.0%, versus 92.0% for content-only MoE and 87.9% for concatenation, with the largest gain under modality drop; the same model also improves mAP50 by +5.2/+5.3 on the natural day/night split. These results suggest that preserving fusion-time reliability as an explicit signal benefits both adaptive fusion and post-fusion conditional computation.
Authors: Jiayu Zhang, Shuo Ye, Jiajian Huang, Yawen Cui, Taorui Wang, Wei Xia, Zeheng Wang, Haowen Tang, Yelin Wang, Hui Ma, Zitong Yu
Abstract: Deception detection is a critical and highly challenging task within affective computing and behavioral analysis. Existing deep learning methods typically treat this task as a straightforward classification problem; however, this black-box approach lacks interpretability and fails to capture the complex logical deduction processes utilized by human experts when identifying lies. While Multimodal Large Language Models (MLLMs) have shown potential, applying them effectively requires a bridge between low-level audiovisual cues and high-level logical reasoning. In this paper, we propose DeceptionX, a novel MLLM framework that shifts the paradigm of deception detection from black-box classification to an interpretable Observe-Think-Summarize reasoning process. To address the scarcity of high-quality reasoning data, we first constructed DeceptChain, a high-quality dataset developed through a human-in-the-loop process. This dataset synthesizes fine-grained visual and auditory evidence (such as micro-expressions and vocal tremors) into structured chain-of-thought reasoning data. Furthermore, we propose a three-stage training pipeline and a Discrepancy-Aware Redundancy Elimination~(DARE) strategy for DeceptionX to further enhance the model's generalization capabilities. Extensive experiments demonstrate that DeceptionX not only outperforms existing MLLM baselines and state-of-the-art methods on standard real-world benchmarks but also provides transparent, expert-level reasoning paths, bridging the critical gap between accuracy and interpretability in multimodal deception detection.
Authors: Armel Yara
Abstract: We introduce HorusEye, Language as Dynamic Attention for Emergency Visual Analysis. Our investigation followed five stages. The first one is benchmarking RefCOCO-Degraded, a dataset of 15,244 images (3,811 base images x 4 conditions: Clean, Fog, Smoke and Thermal) with systematic visual degradation. Through four research questions, we evaluate multiple VLMs (Gemini, Qwen2-VL, BLIP-2, LLaVA, Kosmos-2) across visual grounding the second stage, language feedback recovery the third one, health VQA tasks the fourth, and hallucination analysis the final stage. Our key finding is that language feedback effectiveness is model-dependent: Gemini achieves +47.3% improvement in thermal conditions through iterative language feedback, while Qwen2-VL shows -5.1% degradation under the same protocol. We also identify the 'Thermal Paradox' where cropping strategies that improve RGB performance catastrophically fail in thermal imagery. Furthermore, BLIP-2 uniquely hallucinates more under degradation, making it unsuitable for emergency deployment
Authors: Jiaju Han, Ben Zhang, Xuemeng Sun, Qike Zhang, Yuxian Dong, Dingyi Lu, Chengyin Hu, Luwei Yang, Fengyu Zhang, Yiwei Wei, Jiujiang Guo
Abstract: Remote sensing vision-language models have advanced Earth observation, but available large-scale vision-language resources remain RGB-centered, leaving complementary infrared information underexplored. Infrared observations provide distinctive intensity structures, object boundaries, and illumination-invariant cues that complement conventional RGB imagery, yet large-scale RGB-infrared-text resources remain scarce. We introduce FusionRS, the first large-scale RGB-infrared-style-text dataset for controlled dual-modal remote sensing vision-language learning. It contains 600,000 spatially aligned pairs created by translating diverse public RGB remote sensing images into infrared-style counterparts. Each pair retains a conventional scene caption, and a curated subset adds 45,913 IR-aware captions describing observable intensity, contrast, texture, and structure while preserving scene semantics. We train CLIP-style models for RGB-infrared-style-text alignment and adapt a generative vision-language model with mixed task-conditioned caption supervision. Evaluation covers cross-modal retrieval, scaling and supervision ablations, sensor-captured transfer, and strictly held-out captioning and VQA. FusionRS substantially improves RGB-infrared-style alignment and infrared-to-text retrieval over RGB-only and non-IR-aware settings. Ablations show that IR-aware captions improve task-conditioned infrared description, demonstrating the value of modality-specific supervision. FusionRS provides a scalable foundation for controlled RGB-infrared remote sensing vision-language learning.
Authors: Niyoj Oli, Sachin Acharya, Sandesh Pokhrel, Sanjay Bhandari, Ramesh Rana, Nikesh Mani Shrestha, Ram Bahadur Gurung, Yash Raj Shrestha, Prashnna K Gyawali, Binod Bhattarai
Abstract: Gastrointestinal cancers represent a growing health burden in the South Asian region, driven largely by rapid changes in socio-economic conditions and lifestyle habits. However, early diagnosis remains limited by inadequate equipment, financial resources, and scarce GI expertise. AI-assisted diagnosis and report generation show great promise in alleviating this problem by providing non-specialist healthcare workers the technical expertise to perform diagnosis. Yet, almost all open-source, publicly available datasets are predominantly collected from the European region, with minimal representation from the South Asian region. The lack of open-source GI datasets from diverse geographic regions has made it difficult to assess whether population bias is present in existing models, and to develop geographically inclusive AI tools for automated GI diagnosis. To address this gap, we introduce SAGE: An Expert-Annotated South Asian GI Endoscopy dataset for Multimodal Learning and Hallucination Analysis, for image captioning, multi-label classification, and VQA tasks. It consists of 1,300 images, their captions along with hallucination tag, 18 labels and 14,276 question-answer pairs making it well-suited for diverse range of tasks including classification, benchmarking, and fine-tuning LMMs. We further conducted benchmarking of task-specific models, like multi-class classifiers on the effect of population shift which reveals that such models suffers the most with an average F1 score drop of 54 points on South Asian dataset. Further, benchmarking of contemporary LMMs reveals a substantial drop in the average GREEN score for anatomical landmark detection (0.308) and abnormality detection (0.410). We open-source our dataset under the CC BY-SA 4.0 license, and we hope to encourage others to contribute toward more inclusive dataset representation and help counteract population bias in medical AI.
Authors: Deyin Liu, Jicheng Xu, Lin Yuanbo Wu, Xiaowei Zhao, Xiatian Zhu, Zhe Jin, Anjan Dutta
Abstract: Human image animation, which aims to generate a video of a reference subject following a provided action sequence, has received increasing research interest. With the development of diffusion-based/flow-based video foundation models, existing animation works have began to upgrade the guidance information from 2D skeleton/pose to 3D modeling conditions. Despite achieving reasonable results, these approaches face challenges in synthesizing trajectory-controllable human motion within natural scene under changed camera views. In this work, we present a scene-adaptive human image animation framework that controls both human motion and camera trajectories within a reconstructed 3D environment for video generation. To achieve this, we first develop a ground-adaptive 3D motion retargeting approach to enable user-friendly motion trajectory control adapting to the changes of elevations of ground and orientations automatically. Then we design a viewpoint-adaptive latent fusion mechanism to inject point-cloud geometric priors through scene-visibility masking into the generative process, providing precise guidance of viewpoint changes under camera control. Experiments on two standard human image animation benchmark datasets demonstrate remarkable improvements of our method over the state of the arts in related video generation metics. Project page: https://robinhood256100.github.io/web-disp
Authors: Lianyu Hu, Shengqian Qin, Zeqin Liao, Qing Guo, Liang Wan, Wei Feng, Yang Liu
Abstract: Chain-of-thought (CoT) reasoning has enabled multi-modal large language models (MLLMs) to tackle complex visual reasoning tasks by generating explicit intermediate reasoning steps in natural language. However, this text-based reasoning paradigm is inherently slow at inference time with even thousands of tokens and fundamentally constrained by the expressiveness of natural language. In this paper, we propose CoLT, (Chain of Latent Thoughts), a novel framework that teaches multi-modal models to reason through a chain of latent thought representations instead of verbose text tokens, which can perform thinking with as few as 3 steps. Naively forcing the model to think with latent states easily produces meaningless semantics and makes training unstable. To effectively regulate the latent reasoning process, we introduce a lightweight external decoder that provides step-level supervision for each latent reasoning step in two complementary directions: a forward mode that decodes latent thoughts into the textual reasoning of the next step, and a backward mode that aligns decoder hidden states with the model's latent thoughts given preceding textual context. We further incorporate internal supervision that encourages coherent step-by-step latent transitions. The decoder and internal supervision are removed during inference to maintain high efficiency of latent reasoning. Extensive experiments on eight benchmarks demonstrate that CoLT not only outperforms existing latent reasoning methods such as CODI and SIM-CoT, but also surpasses latent visual reasoning approaches that rely on auxiliary images with costly annotation requirements. Compared to text CoT methods, CoLT can notably reduce the inference time by 10.1$\times$ and text decoding time by 22.6$\times$. Code is released at https://github.com/hulianyuyy/CoLT.
Authors: Ningning Han, Lei Fan, Jia Guo, Yunkang Cao, Xiu Su, Feng Cao, Donglin Di, Tonghua Su
Abstract: The deployment of Industrial Anomaly Detection (IAD) in real-world manufacturing frequently encounters a challenging cold-start bottleneck, in which limited normal samples fail to represent the full normal distribution and only a few anomalies are available. Under such a regime, existing methods struggle to form compact normal boundaries and fail to effectively exploit supervised signals from rare defects. To address this challenge, we propose Anomaly-Rectified Cold-start AD (ArcAD), a plug-and-play calibration framework for reconstruction-based IAD baselines. ArcAD follows a push-pull learning paradigm to construct a compact and discriminative normal boundary under data scarcity. On the one hand, ArcAD projects limited normal samples onto a hypersphere and pulls them into multiple compact clusters to maximize coverage of the normal manifold. On the other hand, it synthesizes pseudo-anomalies on the hypersphere and leverages real anomalies to push the boundary inward and sharpen anomaly discrimination. Extensive experiments on MVTec-AD, VisA, Real-IAD, and MANTA demonstrate that ArcAD significantly outperforms state-of-the-art supervised and unsupervised methods in both single-class and multi-class settings under cold-start conditions. Code is available at: https://github.com/LGC-AD/ArcAD.
Authors: Tianfang Zhang, Lei Li, Chang Liu, Zhenming Peng, Huaping Zhang, Xiangyang Ji
Abstract: Infrared small target detection (IRSTD) aims to identify long distance small targets from complex infrared backgrounds, and is a fundamental task in remote sensing. Deep learning methods have improved IRSTD by learning discriminative image-to-mask mappings, but such feed-forward designs often underuse physical decomposition structure between targets and backgrounds. Deep unfolding methods partially address this issue by embedding model-driven iterations into neural networks, yet existing designs still operate mainly in image domain and use updates and memory mechanisms that are not fully coupled with underlying optimization process. To address these limitations, we propose Latent Consistent Proximal unfolding network (LCPNet). First, we verify that low-rank prior remains valid in latent representations and perform unfolding in this space, preserving physical constraint while avoiding repeated compression of intermediate states. Second, we derive a Latent Consistent Proximal (LCP) solver that evolves each latent variable from its previous state rather than reconstructing through an indirect residual, and stabilizes small target updates through task-adaptive normalization and gain control. Third, we introduce Shared Optimization Memory (SOM), a common historical state shared by all decomposition variables to provide coordinated guidance across unfolding stages. Extensive experiments on four public benchmarks demonstrate that LCPNet achieves accurate and robust detection with low false-alarm rates, together with competitive efficiency among high-accuracy deep unfolding methods. Model and code are available at Model and code are available at https://github.com/Tianfang-Zhang/LCPNet.
Authors: Khang Nhat Hoang Vo, Artem Vazhentsev, Artem Shelmanov, Timothy Baldwin, Yova Kementchedjhieva
Abstract: Vision-language models (VLMs) can recognize entities in clear images yet still fail when answering questions that require factual knowledge beyond what is directly observable. Prior work has either examined individual failure modes in isolation or treated incorrect answers as monolithic, binary failures. We propose a tree-structured framework that organizes failures in knowledge-intensive visual question answering into model-specific operational outcomes. Across two datasets and four VLMs, we observe consistent distributions of operational outcomes: some failures occur before entity recognition, while others persist after the relevant entity is recognized. Visual token representations are most informative for recognition-related decisions. Prompt hidden states predict answer success more effectively, although factual-access attribution remains difficult and exhibits only a weak signal. These pre-generation signals support attribution-guided routing to targeted interventions, including image repair, entity support, question rewriting, and factual evidence.
Authors: Team Moxin, Deyi Ji, Tianrun Chen, Xin Zhang, Jiale Yang, Qi Zhu, An Zhao, Zihao Xie, Han Wang, Xuanyi Liu, Yixiang Zhou, Pei Liu, Yi Tan, Cheng Chen, Dayi Zhu, Mingyu Wei, Hanjie Xu, Jun Liao, Siqi Li, Lingyu Lu, Hongye Fang, Hongming Tan, Youjiang Zhu, Taiyu Zhang, Zejian Li, Chaotao Ding, Zhipeng Liang, Wenxuan Song, Yi Li, Baochuan Yang, Xin Jiang, Ben Feng, Jingyuan Zou, Yanlin Liu, Rong Shi, Lingfeng Li, Liyi Yao, Lanyun Zhu, Yunhe Pan, Lingyun Sun
Abstract: The future of World Models depends not only on scaling model capability, but also on scaling practicality and inference efficiency. High-frame-rate inference enables responsive perception, planning, and control in real-world autonomous systems. To this end, we present MoWorld, a cost-effective yet high-performance Flash World Model with an end-to-end framework spanning data generation, pre-training, distillation, and efficient inference, enabling up to 50 FPS real-time interaction with cinematic visual quality without the need of high-end GPUs. To enable large-scale real-world deployment, MoWorld jointly optimizes model capability and cost throughout the entire development pipeline. Specifically, unlike existing approaches that primarily rely on large-scale video corpora, MoWorld is built upon a scalable 3D-native data engine accumulated from our large-scale 3D vision and generative modeling pipeline, enabling the efficient construction of geometrically consistent training data across diverse real-world and synthetic environments. Based on this foundation, a curriculum cross-frame pre-training strategy for stable and scalable World Model learning, an efficient denoising-step distillation algorithm to reduce diffusion training cost, and a mixed-precision parallel inference framework for low-cost real-time deployment. MoWorld is the first real-time interactive World Model built on the Neural Processing Unit (NPU) and can achieves up to 50 FPS in such the devices, enabling practical and efficient deployment at scale. Comprehensive evaluations demonstrate that MoWorld achieves leading performance; notably, its average inference cost is only 30\%-50\% of that of existing World Models, providing a practical foundation for large-scale real-world applications of World Models. We also demonstrate diverse applications of MoWorld.
Authors: Laiba Khan, Hung-Mao Wu, Wei Lin, Frank Bi, Yousef Abdelhadi, Joshua Jung
Abstract: The increasing accessibility of artificial intelligence has led to a rapid rise in AI-generated videos, making it more difficult to distinguish between real and manipulated content. Many existing detection methods rely on a single model and often struggle to generalize across different types of deepfakes. In this work, we developed a multimodal deepfake detection system that combines both audio and visual analysis using an ensemble of models. The system includes AASIST for audio-based detection, and EfficientNet, XceptionNet, and MesoNet for analyzing visual features in video frames. The pipeline takes a video as input, separates the audio, and extracts face frames using MTCNN. Each model produces a score indicating the likelihood of the input being fake. These scores are then combined using ensemble strategies, including mean averaging and stacking. Mean fusion provides a simple and stable baseline, while stacking uses a trained meta-model to learn how to combine predictions more effectively. Results show that while individual models perform well on the datasets they were trained on, their performance drops when tested on more diverse datasets. The ensemble approach helps improve overall robustness by combining predictions from multiple models, leading to more consistent performance across different types of deepfakes. This suggests that using both audio and visual information together is a more reliable approach for deepfake detection. Our results highlight generalization to unseen manipulations as the central open challenge, with average accuracy around 70%.
Authors: Jiaqi Liu, Haidong Kang, Qihui Zhao, Guo Yu, Jingchao Wang
Abstract: Parameter-efficient fine-tuning enables large language models to adapt to downstream tasks with substantially lower computational and storage cost, and Low-Rank Adaptation (LoRA) is among its most widely used techniques. However, vanilla LoRA assigns a uniform rank to all adapted modules, while existing adaptive methods either incur additional optimization overhead or rely on static weights and local gradients that do not capture task-conditioned representation changes. We propose RSRA, a training-free rank allocator that estimates where adaptation capacity is most needed through forward-only representation sensitivity probing on a small calibration set. Specifically, RSRA uses Spectral Effective Rank to allocate capacity across layers, measures module-wise hidden-state displacement under standardized virtual low-rank updates with the Frechet Distance, and combines both signals through hierarchical normalization to produce a task-aware rank configuration before fine-tuning. Across commonsense reasoning and natural language understanding benchmarks with Qwen3-4B and Mistral-7B, RSRA achieves the highest average performance in all three reported model-benchmark settings and a 1.48x-1.93x speedup in allocation time over the fastest competing pre-allocation method. When integrated with DoRA, LoRA-FA, and PiSSA, RSRA improves 15 of the 18 evaluated combinations and increases the average performance of all three PEFT methods.
Authors: Changfei Fu, Guangcheng Chen, Aoxiang Gu, Haoxiang Liang, Wenjun Xu, Hong Zhang
Abstract: Benefiting from the powerful priors embedded in large-scale pre-training data and the emerging commonsense reasoning ability, large language models (LLMs) have shown unprecedented generalization capabilities in many research fields. Recently, projecting visual embeddings into the language space via vision-language models (VLMs) to achieve sim-toreal and cross-scene generalization has become a prevailing paradigm in the field of Vision-and-Language Navigation in Continuous Environments (VLN-CE). VLN requires an embodied agent to navigate through unseen environments following natural linguistic instructions. We emphasize that a VLN task can be decomposed into a sequence of sub-tasks, each corresponding to a process of 3D spatial interaction with the environments described by instructions such as "walk to the end of the sofa and turn left." However, such spatial interactions involving moving into the image along the direction of depth sensing are puzzling for VLMs as they were predominantly trained on conversations with RGB images. Rather than incorporating depth or 3D geometric information-which VLMs rarely encounter during pretrainingwe propose an alternative approach: fine-tuning VLMs to learn navigation interactions directly in 2D pixel space through autoregressive trajectory generation. Given a linguistic instruction and historical observations, our model sequentially predicts a series of pixel coordinates, drawing a trajectory from the bottom center of the current observation. While prior work has proved that pixel-goal supervision outperforms learning of discrete actions, our experiments further verify that the supervision of pixel-space trajectory significantly enhances VLN performance. Moreover, we demonstrate that our flagship model achieves state-of-the-art level performance with relatively limited computational resources and training data.
Authors: Fatimah Zohra, Chen Zhao, Shuming Liu, Yahya Al Malallah, Bernard Ghanem
Abstract: Gaussian Splatting has emerged as an effective representation for video, but existing methods rely on per-video optimization. This leads to slow encoding and limits generalization across videos. To amortize this optimization, we propose HyperGS, a feedforward, optimization-free approach that directly predicts Gaussian representations from any video in a single forward pass, speeding up encoding and decoding by orders of magnitude while generalizing to out-of-distribution videos at higher resolutions. In HyperGS, we design a factorized spatiotemporal Transformer to extract tokens from video, and a learnable query-based Transformer to obtain 8-parameter Gaussian representations for each video frame. We find that naively predicting Gaussians across diverse videos induces a needle-like degeneration that collapses training, and address this with a rank-based geometric regularizer whose strength adapts dynamically to stabilize optimization. HyperGS achieves encoding at $10^4$--$10^5\times$ the speed of per-video Gaussian optimization at matched reconstruction quality while generalizing zero-shot to $720p$ video, enabling higher-resolution rendering without re-encoding. HyperGS improves PSNR by +2.9--3.1 dB over the prior video encoders on K400, SSv2, and UCF101 at a smaller video representation size. By predicting explicit 2D Gaussians in a single forward pass, HyperGS combines the fast, flexible rendering of Gaussian Splatting with the speed and generalization of feedforward prediction, advancing Gaussians as a practical direction for fast and generalizable video representation.
Authors: Haoyu Fu, Jiafeng Huang, Yuchen Wang, Shengjie Zhao
Abstract: Neural radiance fields and 3D Gaussian Splatting assume that each training image is a sharp and geometrically consistent observation of the scene. Motion blur violates this assumption because a single exposure integrates a continuous range of camera poses. Exposure integration also removes the temporal information needed to recover the corresponding sharp observation. Event cameras preserve this information at microsecond resolution and therefore provide a natural complement to conventional images. Existing event-assisted reconstruction methods predominantly obtain image supervision through analytical inversion of the Event Double Integral. Learned restoration from frames and events offers a second prior. Although weaker when used alone, it fails in different regions and provides complementary evidence. We present JADE-GS, which formulates the combination of these priors as spatial evidence allocation. A lightweight Spatial Prior Router predicts a pixelwise allocation using only the blurry frame and event stream, then fuses the two fixed restorations into an additional supervision target. The router is trained without a sharp reference using consistency with the scene under reconstruction and the measured exposure, and is removed after optimization. Experiments show that JADE-GS achieves leading perceptual quality on both benchmarks, attains the best fidelity on the real benchmark, and remains competitive on the synthetic one. It requires substantially lower training overhead than diffusion-based alternatives and preserves native 3DGS rendering with no generative decoding at inference.
Authors: Xiuyuan Zhu, Ke Lu, Hao Wu, Siwen Jiao, Zijin Du, Dongming Zhang, Jian Xue
Abstract: Visual grounding with multimodal large language models is commonly formulated as autoregressive coordinate generation, where a model outputs bounding-box coordinates as text given an image and a referring-expression prompt. While this interface is simple and compatible with instruction following, it introduces a mismatch between training and evaluation: training optimizes token-level likelihood over coordinate strings, whereas grounding quality is measured by geometric overlap. We propose IoU-PD, an IoU-aware privileged distillation method for coordinate-generating multimodal large language models. IoU-PD uses ground-truth boxes not only as coordinate targets, but also as privileged training-time guidance. During training, the student receives the original image and prompt, while a frozen teacher receives a box-marked image and an augmented prompt that indicates the marked region. The student is trained with a supervised fine-tuning anchor and a privileged distillation loss whose token weights reflect both geometric importance and teacher reliability. At inference time, IoU-PD requires no box overlay, privileged hint, teacher branch, or additional prediction module. Experiments on standard referring-expression grounding benchmarks show consistent region-level improvements over strong coordinate-generating baselines, demonstrating that ground-truth boxes can provide useful privileged guidance beyond serving as coordinate labels.
Authors: Xianghong Fang, Litao Guo, Hengchao Chen, Yuxuan Zhang, XiaofanXia, Dingjie Song, Yexin Liu, Hao Wang, Harry Yang, Qiang Sun, Yuan Yuan
Abstract: The effectiveness of modern visual representation learning and autoregressive models critically depends on vector quantization (VQ), which discretizes continuous feature representations using a learnable codebook. Despite its widespread use, existing VQ methods often suffer from training instability and codebook collapse, arising from gradient mismatch induced by the straight-through estimator and the under-utilization of code vectors. In this work, we show that both issues can be traced to a fundamental mismatch between the distributions of feature vectors and code vectors, leading to inefficient representation and information loss. Building on this observation, we propose a distributional matching framework for vector quantization. We introduce principled criteria for desirable VQ behavior and demonstrate through theoretical analysis and empirical evaluation that aligning feature and code vector distributions provides a unifying mechanism for mitigating training instability and codebook collapse. We instantiate this framework using a Wasserstein-based objective with an efficient closed-form under a mild Gaussian approximation, and further show that a nonparametric alternative based on maximum mean discrepancy yields comparable performance. Extensive experiments on visual tokenization benchmarks support the effectiveness and robustness of the proposed approach.
Authors: Carson Rodrigues
Abstract: Brain-encoding foundation models predict fMRI responses to video, audio and text well enough to win the Algonauts 2025 challenge. We ask whether their predicted responses, obtained with no scanner, are a useful feature lens for a human-behavior task: forecasting short-video memorability. Each clip is projected into TRIBE v2's predicted cortical space and scored by ridge regression against a matched control, the model's own V-JEPA2 visual backbone taken before the brain projection. The answer is dataset-dependent. Within Memento10k (499 clips) the backbone wins (Spearman 0.594 vs 0.544); within VideoMem (820 clips) the brain projection wins (0.415 vs 0.368). Because the claim is that the ordering reverses, we test the reversal itself: the dataset-by-representation interaction is +0.097, 95% CI [+0.032, +0.160], two-sided bootstrap p=0.001, and over 10 cross-validation seeds the datasets separate completely (0/10 seeds favor the brain projection on Memento10k, 10/10 on VideoMem). Cross-dataset transfer inherits the split: Memento10k->VideoMem the brain projection wins (+0.076); the reverse loses heavily (-0.311). The VideoMem advantage is not a sample-size artifact (it survives matched training size and PCA-then-ridge) and not mere compression (a compressed, heavily regularized or transfer-tuned backbone stays below it). Predicted-brain features thus carry a small but real memorability signal the backbone misses on one dataset and not the other: a dataset-specific representation, not a domain-general prior. A vision-orthogonal component (partial Spearman 0.19, permutation p=2.5e-4) localizes to ventral occipito-temporal cortex, and predicted BOLD dynamics add nothing beyond the time-average because 3-4 samples per clip cannot resolve the sub-second late memorability response. Our pre-specified within-dataset hypothesis returned NO-GO; the reversal is what survived.
Authors: Weiwei Tan, Junxian Li, Rui Wang, Zhenhua Xu, Yanjun Zhang, Yu Leo Zhang
Abstract: Unified multimodal models (UMMs) have recently demonstrated powerful instruction-based image editing capabilities, while also raising serious concerns about the unauthorized manipulation of personal portraits. We investigate a novel and practical problem: protecting facial identities against unauthorized editing of UMMs. Existing diffusion-based and VLM-based protection methods often become ineffective because they typically disrupt only a single visual branch. To understand this limitation, we conduct a feature-level analysis of the understanding and generation branches in unified image editing models. Our observations show that the structural agreement between these two branches is closely related to successful image editing. When only one branch is distorted, the model may still recover identity information from the other branch. Based on this, we propose Cross-Branch Conflict as a Shield (CCS), a unified adversarial protection framework. CCS jointly drives the ViT and VAE representations away from their clean counterparts. It also uses a linear Centered Kernel Alignment (CKA) objective to disrupt the structural consistency between the two branches. By degrading reliable identity information in both visual pathways and inducing incompatible cross-branch representations, CCS effectively prevents UMMs from recovering consistent facial identity cues during editing. Extensive experiments suggest that CCS consistently provides stronger protection in suppressing identity-preserving edits. Codes are in the supplementary material.
Authors: Cheng Guo, Qiming Cao, Shengkai Xu, Haoyu Xie, Kaixiang Su, Pu Wang, Hongfei Xue
Abstract: Millimeter-wave (mmWave) radar enables privacy-preserving and illumination-robust human motion reconstruction, but training generalizable models typically requires costly paired radar-motion recordings. Simulation can scale such supervision, yet even physics-based simulators cannot fully reproduce real-world multipath, clutter, hardware-specific response statistics, or distance-dependent resolution degradation, leaving a sim-to-real gap. We present mmSimPrior, a simulation-pretrained framework that factorizes transferable knowledge into signal, motion, and radar-to-motion mapping priors. To learn transferable signal and motion priors, we pretrain a multimodal radar encoder with a physics-informed domain-randomization curriculum designed to mitigate the sim-to-real gap by approximating real-world propagation- and acquisition-level variations, while a joint-temporal tokenizer learns a discrete prior over plausible human motion. A dual-mode mapping module predicts either motion-code distributions for structurally constrained zero-shot reconstruction or continuous motion parameters for flexible adaptation from limited real data. We further construct a 4.2M-frame, 31K-sequence dataset suite and introduce a No-Overlap Setting that prevents any exact subject-environment-location-motion tuple from appearing in both the adaptation and test sets. Experiments on mmSimPrior-Real and RT-Pose demonstrate consistent gains: with only 24 paired real sequences, mmSimPrior-Reg reduces MPJPE by 24.7-39.0% over the strongest baseline across the three environments, while mmSimPrior-Cls reduces zero-shot MPJPE by 8.5% without fine-tuning.
Authors: Jinsen Su, Yongdong Luo, Yuexiao Ma, Yibo Hu, Meiguang Jin, Xiawu Zheng
Abstract: Existing token compression methods for omnimodal large language models typically rely on one modality to determine what to retain in the other. We show that this assumption often breaks down: for the same query, audio and video relevance often peaks at different moments. This cross-modal salience mismatch makes unidirectional guidance prone to discarding answer-critical cues under aggressive compression. We propose OmniScope, a training-free token compression framework that uses the query as a shared semantic anchor while estimating relevance separately for audio and video. OmniScope allocates modality-specific token budgets, prunes visual tokens with an anchor-delta strategy that preserves both global context and temporal changes, and merges audio tokens within each second to reduce redundancy while maintaining temporal continuity. Across four audio-video benchmarks and two Qwen2.5-Omni model scales, OmniScope achieves the best average accuracy across all compression settings. At 25% overall token retention, it delivers up to 3.53x prefill speedup and more than 15% GPU memory reduction, with only a 0.35-point drop in average accuracy. These results suggest a simple design principle for OmniLLM inference: share the query across modalities, but not the salience estimates. The code is available at https://github.com/MAC-AutoML/OmniScope.
Authors: Yuhui Zeng, Wang Chen, Jinfa Huang, Tianyu Xie, Yongdong Luo, Jiayi Ji, Xiawu Zheng, jiebo Luo
Abstract: Existing Large Vision-Language Models (LVLMs) struggle with long-form video understanding due to the quadratic computational cost of visual tokens. While recent efficient methods attempt to compress tokens via hard pruning or uniform merging, they operate strictly in the spatial feature domain, where robust structural context and discriminative semantic details are inherently entangled. In this work, we propose WaveZip, a joint signal-frequency-domain framework for efficient video inference. Driven by the insight that temporal redundancy resides in low-pass approximation scales while spatial saliency strongly correlates with high-frequency components, WaveZip leverages Discrete Wavelet Transforms (DWT) to disentangle these signals. Temporally, it employs 1D DWT to analyze query-frame relevance, and the resulting high-frequency coefficients are further gated by inter-frame differences, with both signals jointly driving the dynamic allocation of a precise frame-level token budget. Spatially, a 2D DWT decomposes features into low-frequency approximations and high-frequency detail components, where the high-frequency coefficients are modulated within query-salient regions to regulate spatial reconstruction. Importantly, WaveZip requires no task-specific training and can be seamlessly integrated into off-the-shelf LVLMs to boost inference efficiency. Extensive experiments on long video understanding benchmarks demonstrate that WaveZip retains 99.6% of the full performance under an extreme 10x compression ratio, consistently outperforming state-of-the-art methods.
Authors: Nao Sugeta, Kaito Shiku, Shinnosuke Matsuo, Ryoma Bise
Abstract: In prostate cancer histopathology, the Gleason Score is determined by the most frequent (Primary) and second most frequent (Secondary) Gleason patterns within a whole-slide image. Although these slide-level labels are routinely available in clinical practice, instance-level Gleason annotations are rarely provided, making patch-level learning challenging. We propose a Multiple Instance Learning (MIL) framework that estimates instance-level Gleason patterns from slide-level Primary and Secondary labels. The proposed method formulates instance-level learning according to the clinical definition of the Gleason Score by aggregating instance predictions into class counts and explicitly modeling the Primary pattern, Secondary pattern, and their dominance. Experimental results demonstrate that the proposed formulation enables effective instance-level learning and outperforms existing MIL approaches on the SICAP-MIL dataset.
Authors: Qizhe Wei, Xianda Guo, Shaocong Xu, Hong Li, Runyi Yang, Hao Zhao
Abstract: Stereo matching and surface normal estimation are fundamental tasks in 3D vision. However, existing feed-forward stereo methods still struggle to produce reliable predictions in challenging regions, mainly due to the lack of strong geometric priors. In this paper, we propose $\textbf{GeoStereo}$, a unified stereo geometry estimation framework that leverages powerful diffusion priors to jointly predict disparity and surface normals. Specifically, GeoStereo couples a feed-forward stereo matching pipeline with a diffusion-based normal estimation branch. To enable effective interaction between the two tasks, we introduce a disparity to normal initialization strategy and construct a warp to left-view condition for the diffusion process. This coupled design allows the diffusion branch to provide strong structural priors that enhance disparity estimation in ill-posed regions, while the feed-forward branch offers reliable geometric guidance for accurate normal prediction. Extensive experiments show that GeoStereo performs reliably in challenging scenarios, including low-light environments, highly reflective surfaces, and transparent objects. Under zero-shot settings, it achieves Rank-1 disparity estimation on multiple benchmarks, including KITTI and NYUv2, and delivers the best normal estimation accuracy on many real indoor benchmarks, such as iBims-1 and ScanNet. Project page: https://qz-wei.github.io/GeoStereo.github.io/
Authors: Kai Wang, Ziheng Ouyang, Xuying Zhang, Ming-Ming Cheng, Qibin Hou
Abstract: With the growth of gaming, animation, and virtual reality industries, the demand for efficient generation of stylized 3D assets is rapidly increasing. However, existing approaches still struggle to jointly preserve style fidelity, geometric consistency, and generation efficiency, as most of them still rely on indirect 2D-to-3D stylization pipelines. This motivates a native 3D stylization framework that can explicitly disentangle style from geometry while remaining efficient. To this end, we propose DreamStyle3D, an efficient framework for stylized 3D asset generation built on a Decoupled Dual Cross-Attention mechanism. Our method explicitly separates geometric and stylistic features to enable efficient style injection while preserving structural consistency, and further adopts a lightweight training strategy to enhance style consistency and model generalization. In addition, we build an automated data pipeline and construct a dataset of about 15K content-style-stylized triplets for training and evaluation. Extensive experiments demonstrate that our DreamStyle3D can generate high-fidelity, geometrically consistent stylized 3D assets within 10 seconds, substantially improving efficiency while maintaining superior style quality and offering a new solution for 3D content creation. The project is available at https://github.com/NK-JittorCV/nk-3D/tree/main/models/DreamStyle3D.
URLs: https://github.com/NK-JittorCV/nk-3D/tree/main/models/DreamStyle3D.
Authors: Behnam Asadi
Abstract: Automated photogrammetric inspection emits metric measurements from reconstructions whose correctness is normally unknown without an external survey. Can a reconstruction estimate its own reliability with no ground truth, and what would such an estimate measure? We formalise a track-leakage-free hold-out protocol: a deterministic image subset is withheld, and each withheld view is re-localised by resection against only those 3D points supported by at least two retained images, so no view is tested against structure it helped create. We evaluate it on five GNSS-referenced captures (four RTK-fixed) across four sites, 13 ETH3D laser-scan scenes, a EuRoC flight, and 30 IMC 2025 scenes. The protocol is computationally well-posed -- good reconstructions score near-perfect self-consistency (median rotation error 0.003 deg) -- but it does not measure accuracy, for a structural rather than statistical reason. It saturates: confidence stays pinned at 1.00 while true error swings 14.1x within a single capture, and holds at 1.00 on survey-grade truth at 3.4 m, 4.3 m, and 1.7 m / 13 deg. It is blind to coherent distortion: corruption that fragments a reconstruction is caught (1.00 -> 0.96), but corruption yielding a single, internally self-consistent, globally distorted model is not -- at three of four captures such models were wrong by 55-106 m at confidence 1.00. On IMC 2025 the same dichotomy appears with no injected degradation: confidence separates failed from successful reconstructions (rho = 0.68) yet ranks nothing among the successful ones (rho = 0.01). A capture-level meta-analysis of the continuous signal is underpowered and sign-unstable (k = 5, 95% CI [-0.45, +0.75]); the negative result does not rest on it. Track-leakage-free hold-out therefore measures internal geometric consistency: a qualitative fragmentation warning, not a substitute for control-point accuracy assessment.
Authors: Jiatong Li, Leo Liang, Linghe Kong, Yulun Zhang
Abstract: Autoregressive video diffusion models enable real-time streaming video generation. However, errors introduced during self-rollout accumulate over long horizons, manifesting as color drift, motion stagnation, and eventual visual collapse. In this paper, we characterize this phenomenon from a frequency-domain perspective: error accumulation appears as a pronounced energy drift in the low-frequency bands. We further investigate the effectiveness of attention sink in the frequency domain, and find that it improves the video quality by alleviating the spectral energy drift to some extent, but cannot fully resolve it. Motivated by the above analysis, we propose FreqForcing, a training-free framework that addresses error accumulation in long-video generation via Spectral Self-Anchoring (SSA). The proposed SSA leverages the low-frequency components of anchor attention to maintain long-horizon visual stability, while preserving dynamic motion through the high-frequency components of local attention. Our FreqForcing extends Self-Forcing pretrained on 5s clips to two-minute generation, achieving 24x extrapolation. Extensive experiments show that FreqForcing outperforms existing training-free methods quantitatively and qualitatively while remaining competitive with representative training-based approaches.
Authors: Kaiyu Li, Zepeng Xin, Zixuan Jiang, Jing Fu, Lanxuan Xue, Lingyu Zhang, Xiangyong Cao
Abstract: Open-vocabulary Earth observation (EO) aims to localize geospatial concepts specified in natural language rather than a fixed label set. Existing benchmarks, however, usually cover narrow category vocabularies or limited query forms. To fill this gap, we introduce OVEarth-Bench, which extends existing evaluation in two directions: category breadth, through broad hierarchical category coverage with positive and negative expressions, and query diversity, through vocabulary, referring, and reasoning queries. The benchmark supports mask and box localization under a unified zero-shot protocol. We evaluate a broad set of general and EO-specific methods. The evaluation reveals that: (1) the performance of current methods remains limited, while broader category coverage yields more stable model rankings; (2) MLLM-based methods achieve the strongest overall performance; and (3) EO-specific methods generally underperform general models and rarely match the strongest methods. These findings provide guidance for future open-vocabulary EO method design and highlight the importance of developing more realistic, diverse, high-quality, and large-scale benchmarks for reliable evaluation. Our data and evaluation package are released at https://earth-insights.github.io/OVEarth-bench.
Authors: Wenjie Zhu, Yabin Zhang, Wenjun Zeng, Lei Zhang
Abstract: Multimodal Large Language Models (MLLMs) have achieved strong performance on a wide range of vision-language tasks, but often fail under imperfect or shifted contexts. A reliable MLLM should refuse truly out-of-context (OOC) questions with subject-level context shifts while still answering shifted in-context (Shifted IC) questions with non-subject context shifts. Existing benchmarks mainly target OOC or visually unanswerable questions, but overlook answerable Shifted IC cases and cover limited OOC shifts. To fill this gap, we present MMOOC, a large-scale benchmark for evaluating refusal and robust answering abilities of MLLMs. MMOOC contains over 41K image-question pairs, including answerable Shifted IC cases and unanswerable OOC cases, spanning three question formats, eight shift types and six visual scenarios, with data quality ensured through MLLM-based filtering and human verification. We evaluate model responses using Accuracy and Refusal Rate, and further introduce an LLM-as-a-Judge metric to assess the correctness of model reasoning. Experiments on diverse MLLMs show that current models still struggle to balance answer-ability and refusal under shifted contexts. We further analyze key failure patterns and show that post-training can improve robustness. MMOOC will be made publicly available.
Authors: Rui Tang, Wentao Yang, Peirong Zhang, Yongxin Shi, Shun Zhang, Huiguo He, Lianwen Jin
Abstract: Scene text spotting requires high-precision alignment between textual recognition and spatial localization. While visual-token grounding has emerged as a promising formulation for Multimodal Large Language Models (MLLMs), the previous multi-patch paradigm often introduces redundant noise and localization ambiguity, particularly for dense or small text instances. To address this, we propose Single-Patch Text Spotting (SPaTS), a vision-centric framework that routes each text instance through a single anchor visual token and then recovers geometry via full-image refinement. To accurately identify this anchor without oracle labels, we introduce Single-Patch Selective Optimization (SPaSO), a reinforcement learning framework that optimizes discrete visual-token selection using patch-level rewards. To further improve representation robustness and localization precision, we introduce Directional Embedding Alignment (DEA) to suppress unstable norm bias by decoupling feature magnitude and direction, and Patch-Enhanced Decoding (PED) to fuse the routed anchor with language semantics and cross-attend over the full-image feature map for geometry-aware boundary regression beyond coordinate-space surrogates. Extensive experiments demonstrate that SPaTS consistently and significantly outperforms both frontier closed-source MLLMs and OCR MLLMs. Code is available at https://github.com/eeNickTang/SPaTS.
Authors: Jiwen Liu, Shujuan Li, Xiaohan Li, Zijie Meng, Xinyue Liu, Yulong Xu, Yan Zhou, Guoxin Zhang
Abstract: Video re-shooting aims to regenerate videos with controllable camera motion and viewpoint. Existing methods rely on explicit 3D priors, which are limited by reconstruction quality and often perform poorly when synthesizing previously unseen regions, or on paired videos with different camera trajectories, whose scarcity hinders generalization. We revisit video re-shooting through text-driven semantic viewpoint specification, enabling control over shot scale, viewing angle, and first-/third-person perspective. To this end, we propose TARS, a 3D-free video re-shooting paradigm. Timestep-wise sensitivity analysis reveals that camera motion is primarily established during high-noise stages, where coarse spatiotemporal structures are formed. Based on this insight, we introduce self-supervised training to learn camera dynamics and fundamental visual representations without paired re-shooting data or 3D reconstruction. Through data scaling and joint textual-camera conditioning, TARS supports robust camera and viewpoint control, plausibly synthesizing regions beyond the source view under large camera motions while enabling reverse-angle re-shooting and perspective switching. Extensive experiments show that TARS provides more accurate and temporally consistent camera control than prior methods. Project Page: https://ymlinfeng.github.io/TARS.github.io/
Authors: Mingkang Dong, Muxin Pu, Jie Li, Bohan Guo, Songruo Chen, Bin Ren, Xu Zheng, Chen Zhao, Tianwen Qian, Mohamed Elhoseiny, Yuqian Fu
Abstract: Streaming video understanding requires models to continuously retain useful visual evidence before future questions are known. Existing approaches primarily manage the growing visual context according to token importance, temporal redundancy, or segment-level relevance, but rarely organize evidence around objects that persist and evolve over time. Thus, in this paper, we introduce ObjectStream, a training-free framework that treats latent objects as memory anchors for streaming video understanding. ObjectStream induces spatially coherent latent objects directly from frozen Video-LLM representations, links them across frames into persistent anchors, and maintains their histories under a bounded memory budget, without requiring external object detectors or segmentation models. Built on these anchors, ObjectStream preserves three complementary forms of evidence: persistent object histories, transient object changes, and recent visual context. This design enables existing Video Large Language Models (Video-LLMs) to reason over object identities, interactions, and state changes while leaving the underlying model unchanged. Extensive experiments on online streaming and offline long-video benchmarks demonstrate both effectiveness and efficiency. In online streaming evaluation, ObjectStream improves Qwen2.5-VL-7B by 10.0 points on OVO-Bench Real-Time Visual Perception, while reducing peak GPU mem-ory and TTFT by approximately 50%. On offline long-video benchmarks, it surpasses the full-token baseline while discarding 82.5% of visual tokens. These results highlight latent objects as a practical and effective organizing principle for compact streaming video memory.
Authors: Weiquan Lin, Yu Deng, Shiyang Liu, Luping Xiao, Xu Tang, Junzhi Yu, Jiaolong Yang, Lei Zhang, Xingyu Chen
Abstract: Hand-object interaction (HOI) modeling remains challenging because it requires joint reasoning about hand articulation, object geometry, contact, semantics, and dynamics under severe visual uncertainty. Foundation models introduce transferable prior knowledge learned from large-scale cross-domain data, offering new ways to address these challenges beyond task-specific data and models. However, the rapidly growing literature remains fragmented, and existing studies typically describe these methods simply as ``using large models'' without systematically characterizing what knowledge is introduced, where it enters the HOI pipeline, or which HOI uncertainty it helps reduce. This survey presents the first systematic review of foundation-model priors for HOI. We organize the literature into six HOI tasks spanning reconstruction and generation. More importantly, we establish a taxonomy of eight foundation-model sub-priors grouped into geometric, semantic, and visual families. Geometric priors encompass shape retrieval, shape reconstruction, and spatial reconstruction; semantic priors include semantic grounding and language reasoning; and visual priors cover visual representation, image generation, and video generation. Based on this taxonomy, we systematically analyze how different priors are represented, injected, and adapted across HOI pipelines and tasks. Beyond how foundation models empower HOI, we further examine how HOI-derived knowledge is used in robot learning, including human-data pretraining, human-to-robot skill transfer, and HOI-to-robot data generation. Finally, we summarize datasets and evaluation protocols, and discuss limitations and future directions toward more generalizable HOI systems. To support long-term progress, we curate a live repository that continuously aggregates emerging methods and benchmarks.
Authors: Heran Wang, Jianing Sun, Xu Jiang, Genwei Ma, Jigang Duan, Xing Zhao
Abstract: In X-ray CT, metallic objects cause beam hardening, photon starvation, and scattering, leading to projection inconsistency, streaks, dark bands, and structural distortions that compromise clinical diagnosis and quantitative analysis. Existing metal artifact reduction (MAR) methods remain limited: optimization-based methods may leave residual artifacts or blur structures, regression networks may generalize poorly across scenarios, and generative models without sample-specific structural guidance and physical constraints may produce anatomically inconsistent structures. Flow Matching learns a continuous-time velocity field that deterministically transports a source distribution to a target distribution, providing a flexible MAR prior. However, standard unconditional Flow Matching does not exploit sample-specific structure, spatially nonuniform metal-induced degradation, or measured projections. To address these limitations, we propose SCMA, a structure-conditioned and metal-aware Flow Matching framework. First, a linear-interpolation-corrected image is fed into the velocity network with the intermediate state as a sample-specific structural condition, guiding inference toward artifact-free CT images while preserving anatomy. Second, time-varying spatial weights from the metal mask and its distance transform are incorporated into the Flow Matching loss to emphasize severe degradation within and around metal regions. Finally, conditional Flow Matching updates alternate with projection-consistency correction during inference, allowing reliable measurements outside metal traces to constrain predictions. Experiments on simulated and real CT data demonstrate that SCMA more effectively suppresses metal artifacts, preserves local anatomical structures, and reduces hallucination-like structures inconsistent with projection measurements than representative MAR methods.
Authors: Riccardo Raciti, Francesco Guarnera, Francesco Rundo, Luca Guarnera, Sebastiano Battiato
Abstract: In medical imaging, generative models are increasingly deployed to synthesize realistic data and augment limited datasets. Unfortunately, while beneficial for privacy-preserving data sharing, these synthesized images can be repurposed for malicious intents, threatening public health through the creation of Medical Deepfakes. To address this threat, we introduce the K-Space Signature (KSS), a novel forensic framework that isolates hardware and generative traces within the spectral domain. By shifting analysis to the frequency domain, the KSS suppresses macroscopic anatomical variance by subtracting an empirical global anatomical prior computed in the Logarithmic Power Spectral Density (Log-PSD) space. To effectively process these globally distributed spectral artifacts without the local spatial bias inherent to Convolutional Neural Networks, we pair the KSS representation with a novel 3D MLP-Mixer architecture equipped with an ArcFace metric-learning head. Extensive experiments on multi-center 3D MRI datasets demonstrate that this combined approach achieves exceptional detection performance, exceeding 0.99 Accuracy and ROC-AUC on multi-generator synthetic datasets. Furthermore, the framework exhibits robust zero-shot generalization, maintaining strong discriminative power (up to 0.93 Accuracy) on independent datasets acquired from entirely unseen scanners. To ensure full reproducibility, the complete source code and pre-trained models will be made publicly available upon acceptance.
Authors: Zhijun Zeng, Yihang Zheng, Youjia Zheng, Yubing Li, Zuoqiang Shi, He Sun
Abstract: Ultrasound Computed Tomography (USCT) provides a radiation-free option for high-resolution clinical imaging. Despite its potential, the computationally intensive Full Waveform Inversion (FWI) required for tissue property reconstruction limits its clinical utility. This paper introduces the Neural Born Series Operator (NBSO), a novel technique designed to speed up wave simulations, thereby facilitating a more efficient USCT image reconstruction process through an NBSO-based FWI pipeline. Thoroughly validated on comprehensive brain and breast datasets, simulated under experimental USCT conditions, the NBSO proves to be accurate and efficient in both forward simulation and image reconstruction. This advancement demonstrates the potential of neural operators in facilitating near real-time USCT reconstruction, making the clinical application of USCT increasingly viable and promising.
Authors: Jiahang Cao, Qiang Zhang, Jingkai Sun, Jiaxu Wang, Hao Cheng, Yulin Li, Jun Ma, Kun Wu, Zhiyuan Xu, Yecheng Shao, Wen Zhao, Gang Han, Yijie Guo, Renjing Xu
Abstract: Diffusion models have been widely employed in the field of 3D manipulation due to their efficient capability to learn distributions, allowing for precise prediction of action trajectories. However, diffusion models typically rely on large parameter UNet backbones as policy networks, which can be challenging to deploy on resource-constrained devices. Recently, the Mamba model has emerged as a promising solution for efficient modeling, offering low computational complexity and strong performance in sequence modeling. In this work, we propose the Mamba Policy, a lighter but stronger policy that reduces the parameter count by over 80% compared to the original policy network while achieving superior performance. Specifically, we introduce the XMamba Block, which effectively integrates input information with conditional features and leverages a combination of Mamba and Attention mechanisms for deep feature extraction. Extensive experiments demonstrate that the Mamba Policy excels on the Adroit, Dexart, and MetaWorld datasets, requiring significantly fewer computational resources. Additionally, we highlight the Mamba Policy's enhanced robustness in long-horizon scenarios compared to baseline methods and explore the performance of various Mamba variants within the Mamba Policy framework. Real-world experiments are also conducted to further validate its effectiveness. Our open-source project page can be found at https://sagecao1125.github.io/mamba_policy/.
Authors: Jiali Cheng, Hadi Amiri
Abstract: Machine Unlearning aims to remove undesired information from trained models without full retraining from scratch. Despite recent progress, the loss landscape and optimization geometry of unlearning are poorly understood. In this paper, we study machine unlearning through the lens of mode connectivity--the phenomenon that independently trained models can often be connected by smooth low-loss paths in parameter space. We introduce {\em mode connectivity in unlearning} (MCU) and evaluate it across a range of settings, including curriculum learning, second-order optimization, and connectivity across different unlearning methods. We find that many unlearned models lie in connected basins with smooth retain/forget behavior, while changes in training dynamics can move solutions into different basins. MCU also reveals that models within the same basin can differ substantially on privacy metrics, and that unlearning progresses nonlinearly from the original model to the unlearned model. In addition, linear connectivity suggests that most approximate unlearning methods are mechanistically distinct from retraining. Finally, MCU-based ensembling can improve generalization and robustness to relearning attacks, and MCU smoothness correlates with unlearning difficulty. To our knowledge, this is the first study of machine unlearning through the lens of mode connectivity.
Authors: Bo-Hsun Chen, Nevindu M. Batagoda, Dan Negrut
Abstract: Generating synthetic images that closely mimic those from real cameras is instrumental in training visual models and enabling end-to-end visuomotor learning. We introduce DiffPhysCam, a differentiable camera simulator designed to support robotics and embodied AI applications by enabling gradient-based optimization in visual perception pipelines. Differentiable rendering also allows inverse reconstruction of real-world scenes as digital twins, facilitating simulation-based robotics training. Existing virtual cameras offer limited control over intrinsic settings, have difficulty capturing optical artifacts, and lack tunable calibration parameters. DiffPhysCam addresses these limitations through a multi-stage pipeline that provides fine-grained control over camera settings, models key optical effects such as defocus blur, and supports calibration with real-world data. It enables both forward rendering for image synthesis and inverse rendering for 3D scene reconstruction, including mesh and material texture optimization. We show that DiffPhysCam enhances robotic perception performance in synthetic image tasks. As an illustrative example, we create a digital twin of a real-world scene using inverse rendering and use it to set up a virtual experiment in a multi-physics simulation, in which we demonstrate navigation of an autonomous ground vehicle using images generated by DiffPhysCam. The code, data, and output results associated with this paper are available online for reproducibility at Chen, Bo-Hsun (2026a,b,c).
Authors: Md Tanveer Hossain Munim
Abstract: Current evaluation metrics for deep learning weather models create a "Statistical Similarity Trap", rewarding blurry predictions while missing rare, high-impact events. We provide quantitative evidence of this trap, showing sophisticated baselines achieve 97.9% correlation yet 0.00 CSI for dangerous convection detection. We introduce DART (Dual Architecture for Regression Tasks), a framework addressing the challenge of transforming coarse atmospheric forecasts into high-resolution satellite brightness temperature fields optimized for extreme convection detection (below 220 K). DART employs dual-decoder architecture with explicit background/extreme decomposition, physically motivated oversampling, and task-specific loss functions. We present four key findings: (1) empirical validation of the Statistical Similarity Trap across multiple sophisticated baselines; (2) the "IVT Paradox", removing Integrated Water Vapor Transport, widely regarded as essential for atmospheric river analysis, improves extreme convection detection by 270%; (3) architectural necessity demonstrated through operational flexibility (DART achieves CSI = 0.273 with bias = 2.52 vs. 6.72 for baselines at equivalent CSI), and (4) real-world validation with the August 2023 Chittagong flooding disaster as a case study. To our knowledge, this is the first work to systematically address this hybrid conversion-segmentation-downscaling task, with no direct prior benchmarks identified in existing literature. Our validation against diverse statistical and deep learning baselines sufficiently demonstrates DART's specialized design. The framework enables precise operational calibration through beta-tuning, trains in under 10 minutes on standard hardware, and integrates seamlessly with existing meteorological workflows, demonstrating a pathway toward trustworthy AI for extreme weather preparedness.
Authors: Weikai Lin, Haoxiang Li, Yuhao Zhu
Abstract: Hair simulation and rendering are challenging due to complex strand dynamics, diverse material properties, and intricate light-hair interactions. Recent video diffusion models can generate high-quality videos, but they lack fine-grained control over hair dynamics. We present ControlHair, a hybrid framework that integrates a physics simulator with conditional video diffusion to enable precise and controllable dynamic hair rendering. ControlHair adopts a three-stage pipeline: it first encodes physics conditions into per-frame geometry using a simulator, then extracts per-frame control signals, and finally feeds control signals into a video diffusion model to generate videos with desired hair dynamics. This cascaded design decouples physics reasoning from video generation, supports diverse physics, and makes training the video diffusion model easy. Trained on a curated 10K video dataset, ControlHair outperforms text- and pose-conditioned baselines, delivering precisely controlled hair dynamics. We also demonstrate three use cases of ControlHair, including dynamic hairstyle try-on, bullet-time effects, and cinemagraphic. Project page: https://linwk20.github.io/controlhair-web.
Authors: Lorenzo Shaikewitz, Charis Georgiou, Luca Carlone
Abstract: Quantifying the uncertainty of an object's pose estimate is essential for robust control and planning. Although pose estimation is a well-studied robotics problem, attaching statistically rigorous uncertainty is not well understood without strict distributional assumptions. We develop distribution-free pose uncertainty bounds about a given pose estimate in the monocular setting. Our pose uncertainty only requires high probability noise bounds on pixel detections of 2D semantic keypoints on a known object. This noise model induces an implicit, non-convex set of pose uncertainty constraints. Our key contribution is SLUE (S-Lemma Uncertainty Estimation), a convex program to reduce this set to a single ellipsoidal uncertainty bound that is guaranteed to contain the true object pose with high probability. SLUE solves a relaxation of the minimum volume bounding ellipsoid problem inspired by the celebrated S-lemma. It requires no initial guess of the bound's shape or size and is guaranteed to contain the true object pose with high probability. For tighter uncertainty bounds at the same confidence, we extend SLUE to a sum-of-squares relaxation hierarchy which is guaranteed to converge to the minimum volume ellipsoidal uncertainty bound for a given set of keypoint constraints. We show this pose uncertainty bound can easily be projected to independent translation and axis-angle orientation bounds. We evaluate SLUE on two pose estimation datasets and a real-world drone tracking scenario. Compared to prior work, SLUE generates substantially smaller translation bounds and competitive orientation bounds. We release code at https://github.com/MIT-SPARK/PoseUncertaintySets.
Authors: Jingli Liu, Huannan Zheng, Bohao Zou, Kezhou Yang
Abstract: While the unsustainable energy cost of artificial intelligence necessitates physics-driven computing, its performance superiority over full-precision GPUs remains a challenge. We bridge this gap by repurposing the Joule-heating relaxation dynamics of magnetic tunnel junctions, conventionally suppressed as noise, into neuronal intrinsic plasticity, realizing working memory with human-like features. Traditional AI utilizes energy-intensive digital memory that accumulates historical noise in dynamic environments. Conversely, our Intrinsic Plasticity Network (IPNet) leverages thermodynamic dissipation as a temporal filter. We provide direct system-level evidence that this physics-driven memory yields an 18x error reduction compared to spatiotemporal convolutional models in dynamic vision tasks, reducing memory-energy overhead by >90,000x. In autonomous driving, IPNet reduces prediction errors by 12.4% versus recurrent networks. This establishes a neuromorphic paradigm that shatters efficiency limits and surpasses conventional algorithmic performance.
Authors: Dingkun Liu, Yuheng Chen, Zhu Chen, Zhenyao Cui, Yaozhi Wen, Jiayu An, Jingwei Luo, Dongrui Wu
Abstract: Electroencephalography (EEG) foundation models (FMs) have recently emerged as a promising paradigm for brain-computer interfaces, aiming to learn transferable neural representations from large-scale heterogeneous recordings. Despite rapid progress, a fair and comprehensive comparison of existing EEG FMs is still lacking, owing to inconsistent pre-training objectives, preprocessing choices, and downstream evaluation protocols. To fill this gap, we present EEG-FM-Compass. We first review 55 representative models and organize their design choices into a unified taxonomic framework including data standardization, model architectures, and self-supervised pre-training strategies. We then evaluate 12 open source FMs and competitive specialist baselines across 13 EEG datasets spanning nine brain-computer interface paradigms. Emphasizing real-world deployments, we consider both cross-subject generalization under a leave-one-subject-out protocol and rapid calibration under a within-subject few-shot setting. We further compare full-parameter fine-tuning with linear probing to assess the transferability of pre-trained representations, and examine the relationship between model scale and downstream performance. Our results indicate that: 1) linear probing is frequently insufficient; 2) specialist models trained from scratch remain competitive across many tasks; and 3) larger FMs do not necessarily yield better generalization performance under current data regimes and training practices.
Authors: Jincheng Wang, Liwei Luo, Xiaoxuan Tang, Jingxuan Xu, Sheng Zhou, Dajun Chen, Wei Jiang, Yong Li
Abstract: Large language models (LLMs) have recently been applied to Automated Program Repair (APR), yet most existing approaches remain unimodal and fail to use diagnostic signals contained in visual artifacts such as screenshots and control-flow graphs. In practice, many bug reports convey critical information visually (e.g., layout breakage or missing widgets), but directly using such dense visual inputs often causes context loss and noise, making it difficult for MLLMs to ground visual observations into precise fault localization and executable patches. To bridge this semantic gap, we propose \textbf{SVRepair}, a multimodal APR framework with Structured Visual Representation (SVR). SVRepair first fine-tunes a vision-language model, SVR, to uniformly transform heterogeneous visual artifacts into a \emph{semantic scene graph} that captures GUI elements and their structural relations (e.g., hierarchy), providing normalized, code-relevant context for downstream repair. Building on the graph, SVRepair drives a coding agent to localize faults and synthesize patches, and further introduces an iterative visual-artifact segmentation strategy that progressively narrows the input to bug-centered regions to suppress irrelevant context and reduce hallucinations. Across primary repository-level APR benchmarks, SVRepair resolves \textbf{186/517} SWE-Bench M instances (\textbf{35.98\%} over all instances; \textbf{36.47\%} over submitted runs) and \textbf{4/19} visual OmniGIRL instances (\textbf{21.05\%}). On supplementary structured multimodal code reasoning benchmarks, SVRepair reaches \textbf{38.02\%} on MMCode and \textbf{95.73\%} on CodeVision. Code is available at https://github.com/codefuse-ai/CodeFuse-SVR.
Authors: Sining Ang, Yuguang Yang, Chenxu Dang, Canyu Chen, Cheng Chi, Haiyan Liu, Xuanyao Mao, Jason Bao, Xuliang, Bingchuan Sun, Yan Wang
Abstract: Vision-Language-Action (VLA) driving augments end-to-end (E2E) planning with language-enabled visual backbones, yet it remains unclear how vision-language models (VLMs) differ from standard vision-only encoders and whether these differences survive policy learning. We study this question under a unified VLM-hidden + diffusion-policy paradigm, comparing multiple VLM families/scales (e.g., InternVL3 and Qwen3VL) with vision-only encoders (e.g., ResNet, ViT, and EVA-CLIP). We ask three questions: how similar are their representations, do residual differences induce meaningful behavioral differences, and how can they improve accuracy-cost trade-offs? We find that VLM and vision-only policies share a substantial common subspace after policy learning, yet both retain non-transferable residual subspaces. Using a Shared-Unique SAE, we show that these residual factors are behaviorally relevant: vision-only encoders are stronger in simple, geometry-dominant scenarios, whereas VLMs are stronger in long-tail, semantically complex, and interaction-heavy cases. The two policy families also exhibit distinct driving styles, with vision-only models being more conservative on average and VLMs more assertive. Exploiting the complementarity between a VLM branch and a ViT branch yields an oracle upper bound of 93.58 PDMS on NAVSIM. We introduce HybridDriveVLA, which runs both branches and uses a learned trajectory scorer for selection, improving PDMS to 92.10 (+1.30 over the VLM baseline), and DualDriveVLA, a fast-slow variant that invokes the VLM in only 15% of scenarios, achieving 91.00 PDMS (+0.20) with about a 1.9x latency speedup over the VLM baseline. Code will be released.
Authors: Zhiying Jiang, Raihan Seraj, Marcos Villagra, Bidhan Roy
Abstract: Training frontier-scale diffusion models often requires substantial computational resources concentrated in tightly-coupled clusters, limiting participation to well-resourced institutions. While Decentralized Diffusion Models (DDM) enable training multiple experts in isolation, existing approaches require 1176 GPU-days and homogeneous training objectives across all experts. We present an efficient framework that dramatically reduces resource requirements while supporting heterogeneous training objectives. Our approach combines three key contributions: (1) a heterogeneous decentralized training paradigm that allows experts to use different objectives (DDPM and Flow Matching), unified at inference time without any retraining; (2) pretrained checkpoint conversion from ImageNet-DDPM to Flow Matching objectives, accelerating convergence and enabling initialization without objective-specific pretraining; and (3) PixArt-$\alpha$'s efficient AdaLN-Single architecture, reducing parameters while maintaining quality. Experiments on LAION-Aesthetics show that, relative to the training scale reported for prior DDM work, our approach reduces the compute by 16$\times$ and data by 14$\times$. Under aligned inference settings, our heterogeneous configuration achieves better FID and higher intra-prompt diversity than the homogeneous baseline. By eliminating synchronization requirements and enabling mixed DDPM/FM objectives, our framework makes decentralized generative model training accessible to contributors with single GPUs requiring only 24--48GB VRAM.
Authors: Zesheng Liu, Maryam Rahnemoonfar
Abstract: Subsurface stratigraphy contains important spatio-temporal information about accumulation, deformation, and layer formation in polar ice sheets. In particular, variations in internal ice layer thickness provide valuable constraints for snow mass balance estimation and projections of ice sheet change. Although radar sensors can capture these layered structures as depth-resolved radargrams, convolutional neural networks applied directly to radar images are often sensitive to speckle noise and acquisition artifacts. In addition, purely data-driven methods may underuse physical knowledge, leading to unrealistic thickness estimates under spatial or temporal extrapolation. To address these challenges, we develop K-STEMIT, a novel knowledge-informed, efficient, multi-branch spatio-temporal graph neural network that combines a geometric framework for spatial learning with temporal convolution to capture temporal dynamics, and incorporates physical data synchronized from the Model Atmospheric Regional physical weather model. An adaptive feature fusion strategy is employed to dynamically combine features learned from different branches. Extensive experiments have been conducted to compare K-STEMIT against current state-of-the-art methods in both knowledge-informed and non-knowledge-informed settings, as well as other existing methods. Results show that K-STEMIT consistently achieves the highest accuracy while maintaining near-optimal efficiency. Most notably, incorporating adaptive feature fusion and physical priors reduces the root mean-squared error by 21.01% with negligible additional cost compared to its conventional multi-branch variants. Additionally, our proposed K-STEMIT achieves consistently lower per-year relative MAE, enabling reliable, continuous spatiotemporal assessment of snow accumulation variability across large spatial regions.
Authors: Navid Azimi, Aditya Prakash, Yao Wang, Li Xiong
Abstract: Deep neural networks remain highly vulnerable to adversarial perturbations, limiting their reliability in security- and safety-critical applications. To address this challenge, we introduce QShield, a modular hybrid quantum-classical neural network (HQCNN) architecture designed to enhance the adversarial robustness of classical deep learning models. QShield integrates a conventional convolutional neural network (CNN) backbone for feature extraction with a quantum processing module that encodes the extracted features into quantum states, applies structured entanglement operations under realistic noise models, and outputs a hybrid prediction through a dynamically weighted fusion mechanism implemented via a lightweight multilayer perceptron (MLP). We systematically evaluate both classical and hybrid quantum-classical models on the MNIST, OrganAMNIST, and CIFAR-10 datasets, using a comprehensive set of robustness, efficiency, and computational performance metrics. Our results demonstrate that classical models are highly vulnerable to adversarial attacks, whereas the proposed hybrid models with entanglement patterns maintain high predictive accuracy while substantially reducing attack success rates across a wide range of adversarial attacks. Furthermore, the proposed hybrid architecture significantly increased the computational cost required to generate adversarial examples, thereby introducing an additional layer of defense. These findings indicate that the proposed modular hybrid architecture achieves a practical balance between predictive accuracy and adversarial robustness, positioning it as a promising approach for secure and reliable machine learning in sensitive and safety-critical applications.
Authors: Aya Elgebaly, Joris Fournel, Benjamin Laine J{\o}nch Jurgensen, Kamil Mikolaj, Anders Christensen, Martin Tolsgaard, Claes Ladefoged, Aasa Feragen
Abstract: Fairness studies of medical imaging AI often explain subgroup performance gaps through under-representation in the training data. We show that intersectional analysis can disentangle fairness and performance gaps arising from clinical and acquisition confounders that co-vary with the target. As a case, we study scan-time fetal weight estimation from obstetric ultrasound, analyzing two models: a state-of-the-art deep learning (DL) model and the clinical gold-standard Hadlock formula. Using unsupervised slice discovery, we find that high-error subgroups share extreme in the image-acquisition pixel spacing (PS) and in the scan-to-delivery (STD) interval. Of these, PS is an acquisition parameter that can be optimized, while STD is a potential confounder for both PS and our bias diagnostics. Subgroup inspection alone cannot separate them. We disentangle the factors using a model-agnostic analysis with identical metadata partitions and partial regression. Holding STD fixed, the apparent PS effect collapses to a small residual (standardized coefficient $\beta=-0.17$), whereas holding PS fixed, STD dominates error ($\beta=+0.56$). Both models degrade with increasing STD, including the biometric formula, indicating much of the error is intrinsic to the prediction target rather than imaging. The DL model is $\sim$1.5$\times$ more sensitive to STD than Hadlock, though it remains more accurate in every subgroup. We conclude that fairness analyses need to carefully analyze potential confounds, or risk attributing an effect such as temporal or acquisition-related dependency to demographics.
Authors: Chaymae Yahyati, Ismail Lamaakal, Khalid El Makkaoui, Ibrahim Ouahbi
Abstract: Model quantization is widely used to reduce memory, latency, and deployment cost, and is typically judged by whether predictive accuracy is preserved. In decision systems that provide algorithmic recourse, however, accuracy preservation is not sufficient: a small actionable change that flips the decision of a full-precision model may fail after quantization, or require a substantially larger intervention. This paper studies this deployment mismatch and introduces counterfactual sensitivity under quantization, a framework for measuring how compression changes recourse behavior. We propose two metrics: Validity Drop (VD), which measures the fraction of full-precision recourse actions that no longer achieve the target outcome after quantization, and Counterfactual Recourse Gap (CRG), which measures the increase in minimal recourse cost under the quantized model. To mitigate this failure mode, we introduce Counterfactual-Faithful Quantization (CFQ), a quantization-aware training method that jointly learns quantizer parameters and mixed-precision bit allocation while preserving the target prediction at teacher-generated recourse points. CFQ is compatible with standard LSQ/PACT-style quantizers and mixed-precision policies, and can also be instantiated as a training-free calibration procedure for post-training quantization. Experiments on Adult, German Credit, and COMPAS show that standard QAT and mixed-precision baselines can preserve accuracy while substantially degrading recourse stability. At matched accuracy and bit budget, CFQ consistently reduces VD and CRG; for example, on Adult, CFQ reduces VD/CRG from $0.121/0.162$ for an accuracy-centric mixed-precision baseline to $0.061/0.071$.
Authors: Jan Nausner, Michael Hubner
Abstract: Heterogeneous sensor fusion is vital for detecting, localizing, and classifying CBRNE threats. However, individual sensors are often only capable of detecting a subset of relevant threats with varying reliability or can even provide only indirect threat indications, making threat classification challenging. Furthermore, high clutter rates on the sensor side present a great challenge for fusion systems. Additionally, the limited availability of high quality datasets hinders the advancement of learning-based detection and classification models in smart sensors. To mitigate these sensor related shortcomings, a context-aware and domain knowledge-enhanced fusion process is proposed. First, a novel evidence hierarchy is established that enables modeling of direct, indicative, and contextual information. Second, contextual information about the environment is introduced into the fusion process, by collecting, processing, and exploiting OSINT inputs. Third, all levels of the evidence hierarchy are used to craft a Bayesian threat type classification mechanism with domain knowledge-informed priors. The proposed methodology is evaluated in simulated scenarios, and the results demonstrate the benefit of the proposed fusion approach in terms of robustness to clutter and prior mismatch, with an overall classification accuracy of up to 95%.
Authors: Pratheswaran Hariharan, Haiping Xu, Donghui Yan
Abstract: Multimodal large language models (MLLMs) have demonstrated strong capabilities in vision-language understanding and natural-language response generation. However, these systems can still produce overconfident predictions and hallucination-like outputs, particularly when the visual evidence is weak, ambiguous, or semantically inconsistent. Most existing approaches focus on improving multimodal representation alignment or retrieval-augmented generation, while providing limited mechanisms to quantify instance-level prediction reliability or identify incorrect visual outputs. This work proposes a retrieval-augmented reliability-aware inference framework for trustworthy multimodal visual understanding. The proposed framework constructs an external visual evidence database using pretrained visual embeddings and nearest-neighbor retrieval over normalized feature representations. Retrieved evidence is used to estimate prediction trustworthiness through multiple reliability indicators, including similarity strength, class-support agreement, evidence margin, entropy-based uncertainty, and an aggregate reliability score. Based on these signals, a decision gate determines whether the system should accept the prediction, answer with caution, or abstain/fallback when evidence is insufficient. A multimodal response-generation layer then produces a final user-facing response conditioned on the reliability decision. Experiments on ImageNet-100 demonstrate that the proposed reliability-aware framework improves accepted prediction accuracy from 85.84\% to 88.88\% at 89.04\% coverage. The hallucination-like accepted wrong-answer rate is reduced from 14.16\% to 11.12\%. These results show that integrating retrieval evidence, reliability estimation, and selective decision gating can improve calibration and reduce overconfident visual errors without retraining large multimodal models.
Authors: Kwan Soo Shin
Abstract: AI in radiology and other safety-critical workflows is evaluated on the hazards it is told to find, yet harm arises disproportionately from hazards no one specified. We show that conditioning a language or vision model on a narrow task suppresses its reporting of co-present, safety-critical signals it can otherwise report, a behavioral analogue of human inattentional blindness. Across radiology text scenarios and thoracic-image vision tasks, ordinary focused instructions suppressed reporting by up to 0.92; the gap ranged from minimal to complete across seven models, did not vary monotonically with scale, and persisted in a reasoning model, while one flagship model showed a robust safety-reporting override. We term this dissociation the Inattentional Gap: a system can score near-perfectly on specified hazards while omitting co-present safety-critical hazards. In a 24-scenario probe, an independent open-ended critic restored every omitted finding. We propose reporting-complete evaluation as an admission criterion for safety-critical deployment.
Authors: Ryuji Oi, Hikari Otsuka, Kosuke Matsushima, Yuki Ichikawa, Masato Motomura, Tatsuya Kaneko, Daichi Fujiki
Abstract: Vision-Language-Action (VLA) models have emerged as a promising approach for generalizable robotic manipulations. In particular, flow-matching-based VLA models have shown remarkable success due to their capability to generate precise and smooth action sequences and capture multimodal distributions. However, the iterative denoising process in the action head acts as a major computational bottleneck, posing a critical challenge for real-time deployment. To address this challenge, we propose ActionCache, a plug-and-play external cache that opportunistically reuses past intermediate actions to warm-start generations from the vicinity of target actions, drastically reducing the inference latency. Specifically, ActionCache stores the intermediate actions with compact multimodal keys, which enables retrieval from similar past contexts across different episodes or even different tasks. Experimental results in simulation and real-world environments demonstrate that ActionCache maintains high task success rates in a low-latency regime, achieving action head inference acceleration of up to $10.44\times$ and $40.17\times$ for representative flow-based VLA, $\pi_{0.5}$ and GR00T-N1.6, respectively.
Authors: Binghui Chen, Pandeng Li, Yu Liu, Jingren Zhou
Abstract: Music generation foundation models have recently attracted significant industry attention. However, achieving efficient generation and high-fidelity long-form audio while supporting controllability remains challenging. To address these needs, we present \textbf{WanSong}, a simple yet powerful approach for long-form, commercial-grade song generation. Unlike autoregressive (AR) and cascaded multi-stage pipelines (\eg, AR followed by diffusion), \textbf{WanSong} is a pure diffusion-based model that directly generates high-fidelity, multilingual songs up to 5 minutes and outputs dual stems (vocals and background music) in a single run. In addition, our diffusion framework enables faster inference through step-distillation, and offers an efficient pathway for fine-tuning and customization to support downstream editing tasks.
Authors: Zhiwei Wei, Yonghe Sun, Zhenjia Liu, Wenjia Xu, Chao He, Weihua Dong, Chunbo Liu, Hua Liao
Abstract: Spatial understanding is crucial for foundation models (FMs), and maps have long helped humans organize and reason about geographic information. This study examines whether choropleth maps remain useful for machine spatial understanding when models can directly process structured geodata. We introduce ChoroplethMap-Bench, a controlled benchmark containing 2,400 synthetic choropleth maps, corresponding GeoJSON data, and 12,000 questions across five cognitive dimensions: Identify, Spatial Recognition, Compare, Rank, and Delineate. We evaluate 22 open-source and proprietary models under three input conditions: Data Only, Map Only, and Data + Map. The results show that maps substantially improve spatial reasoning, especially when combined with symbolic data and for tasks requiring higher-level understanding of spatial patterns. We further analyze the effects of map type, color hue, and spatial structure, as well as prompting strategies, language, geographic context, decoding settings, classification methods, and response stability. Overall, the Data + Map condition achieves the strongest performance, demonstrating that maps remain valuable external representations for foundation model spatial reasoning.
Authors: Shiyong Meng, Bolei Chen, Ping Zhong, Yang Wan, Rongzhi Wang, Jiazhi Xia, Jianxin Wang
Abstract: Visual Floorplan Localization (FLoc) has emerged as a promising solution for indoor localization by matching egocentric images against minimalist structural maps. However, due to cross-modal information asymmetry and repetitive indoor layouts, visual FLoc is fundamentally challenged by multimodal pose distributions, where visually identical observations map to distinct, spatially separated locations. Existing ray-matching-based methods tackle this by explicitly predicting sparse geometric or semantic rays, which inherently incur information loss and demand resource-intensive preprocessing alongside exhaustive matching during inference. In this paper, we bypass the intermediate ray-matching paradigm and propose a coarse-to-fine visual FLoc framework that progresses from uncertainty to determinism. In the coarse stage, we design an image-conditioned pose diffusion model to parameterize the continuous multimodal pose distribution, effectively routing stochastically initialized pose particles toward distinct candidate modes. In the refinement stage, we propose a localized refiner that predicts bounded sub-meter pose residuals from candidate-centered floorplan crops, where structural ambiguities are largely eliminated. Our method effectively balances global multi-hypothesis tracking and local sub-meter refinement without requiring any offline map preprocessing or test-time lookup tables. Comprehensive results on the S3D (full) and ZInD benchmarks demonstrate that our approach achieves state-of-the-art accuracy and robustness.
Authors: V. S. Usatyuk, D. A. Sapozhnikov, S. I. Egorov
Abstract: We propose Kohn-Sham Spectral Embedding (KSSE), an energy-based model replacing the dense classifier of convolutional neural networks with a sparse-graph spectral embedding evaluated at the Nishimori temperature of an associated Random-Bond Ising Model (RBIM). Mapping pre-trained features onto quasi-cyclic low-density parity-check graphs with a regularized Laplacian acting as a Kohn-Sham Hamiltonian decomposes the system into D independent single-channel spectral problems. These are solved in O(N log N + k_mode^2 N) time via the Fast Fourier Transform on circulant blocks a consequence of Pontryagin self-duality\^a with low-mode Rayleigh-Ritz refinement. Instead of eliminating all frustrated cycles, graph topology is optimized via star-domain surgery, using edge shifts to enforce certified local convexity around codewords while bounding residual frustration. Multi-scale fractal analysis and the learning-rate landscape certify the transition from rough landscapes to star-domain basins. Our rigorous theoretical framework establishes: a generalized Ihara-Bass identity linking belief propagation to the regularized Laplacian; a non-backtracking growth trichotomy where frustration enters as a gauge-invariant Z_2 flux; a trapping-set spectral test; an even-subgraph partition function expansion; exact additive separability with a cup-product obstruction; and a loop-series exchange-correlation bound certifying sub-percent factorization error at girth >= 6. Evaluated on ImageNet-1000 with frozen EfficientNet-B4 features under a transductive protocol, KSSE achieves 88.93% Top-1 accuracy using ~21.24M parameters, outperforming Swin-L (197M, 86.4-87.3%) and matching ViT-H/14 (632M, 88.0-89.5%) while reducing model size by 10x and 30x, respectively.
Authors: Sebastian Doerrich, Daniel W\"urtinger, Francesco Di Salvo, Shyam Nandan Rai, Christian Ledig
Abstract: Adapting deep learning models to profound clinical heterogeneity typically relies on parameter-efficient fine-tuning (PEFT) to avoid the severe overfitting associated with full end-to-end network updates. Although PEFT successfully navigates limited data scenarios, it inherently forces the training of a separate, isolated adapter for every specific diagnostic task. Consolidating these isolated adapters into a single generalist network risks negative transfer, as optimization gradients from conflicting visual domains interfere. To address this, we propose MoPET, a mixture-of-experts (MoE) method that uses a learned sparse router to direct each input through a small subset of low-rank PEFT experts injected into a frozen foundation model, sharing capacity across datasets while limiting cross-domain gradient conflict. Through selected evaluations on the MedMNIST benchmark, we first establish that PEFT outperforms full network updates, improving average accuracy from 86.50% to 88.97%. We then show that a single MoPET model consolidates four heterogeneous datasets into one network, improving average accuracy over the best isolated PEFT adapters (93.46% versus 92.83%). Finally, we show that co-training with auxiliary datasets improves accuracy on data-constrained clinical targets, raising average target accuracy over the strongest isolated adapter from 81.58% to 83.58%. Our source code is publicly available at https://github.com/sdoerrich97/mopet.