Authors: Jingkang Yang, Shuai Liu, Hongming Guo, Yuhao Dong, Xiamengwei Zhang, Sicheng Zhang, Pengyun Wang, Zitang Zhou, Binzhu Xie, Ziyue Wang, Bei Ouyang, Zhengyu Lin, Marco Cominelli, Zhongang Cai, Yuanhan Zhang, Peiyuan Zhang, Fangzhou Hong, Joerg Widmer, Francesco Gringoli, Lei Yang, Bo Li, Ziwei Liu
Abstract: We introduce EgoLife, a project to develop an egocentric life assistant that accompanies and enhances personal efficiency through AI-powered wearable glasses. To lay the foundation for this assistant, we conducted a comprehensive data collection study where six participants lived together for one week, continuously recording their daily activities - including discussions, shopping, cooking, socializing, and entertainment - using AI glasses for multimodal egocentric video capture, along with synchronized third-person-view video references. This effort resulted in the EgoLife Dataset, a comprehensive 300-hour egocentric, interpersonal, multiview, and multimodal daily life dataset with intensive annotation. Leveraging this dataset, we introduce EgoLifeQA, a suite of long-context, life-oriented question-answering tasks designed to provide meaningful assistance in daily life by addressing practical questions such as recalling past relevant events, monitoring health habits, and offering personalized recommendations. To address the key technical challenges of (1) developing robust visual-audio models for egocentric data, (2) enabling identity recognition, and (3) facilitating long-context question answering over extensive temporal information, we introduce EgoButler, an integrated system comprising EgoGPT and EgoRAG. EgoGPT is an omni-modal model trained on egocentric datasets, achieving state-of-the-art performance on egocentric video understanding. EgoRAG is a retrieval-based component that supports answering ultra-long-context questions. Our experimental studies verify their working mechanisms and reveal critical factors and bottlenecks, guiding future improvements. By releasing our datasets, models, and benchmarks, we aim to stimulate further research in egocentric AI assistants.
Authors: Eliza Kosoy, Annya Dahmani, Andrew K. Lampinen, Iulia M. Comsa, Soojin Jeong, Ishita Dasgupta, Kelsey Allen
Abstract: Understanding geometry relies heavily on vision. In this work, we evaluate whether state-of-the-art vision language models (VLMs) can understand simple geometric concepts. We use a paradigm from cognitive science that isolates visual understanding of simple geometry from the many other capabilities it is often conflated with such as reasoning and world knowledge. We compare model performance with human adults from the USA, as well as with prior research on human adults without formal education from an Amazonian indigenous group. We find that VLMs consistently underperform both groups of human adults, although they succeed with some concepts more than others. We also find that VLM geometric understanding is more brittle than human understanding, and is not robust when tasks require mental rotation. This work highlights interesting differences in the origin of geometric understanding in humans and machines -- e.g. from printed materials used in formal education vs. interactions with the physical world or a combination of the two -- and a small step toward understanding these differences.
Authors: Brian Pulfer, Yury Belousov, Vitaliy Kinakh, Teddy Furon, Slava Voloshynovskiy
Abstract: The study of security in machine learning mainly focuses on downstream task-specific attacks, where the adversarial example is obtained by optimizing a loss function specific to the downstream task. At the same time, it has become standard practice for machine learning practitioners to adopt publicly available pre-trained vision foundation models, effectively sharing a common backbone architecture across a multitude of applications such as classification, segmentation, depth estimation, retrieval, question-answering and more. The study of attacks on such foundation models and their impact to multiple downstream tasks remains vastly unexplored. This work proposes a general framework that forges task-agnostic adversarial examples by maximally disrupting the feature representation obtained with foundation models. We extensively evaluate the security of the feature representations obtained by popular vision foundation models by measuring the impact of this attack on multiple downstream tasks and its transferability between models.
Authors: Daniel C. Moura, Shizhan Zhu, Orly Zvitia
Abstract: This paper presents the Nexar Dashcam Collision Prediction Dataset and Challenge, designed to support research in traffic event analysis, collision prediction, and autonomous vehicle safety. The dataset consists of 1,500 annotated video clips, each approximately 40 seconds long, capturing a diverse range of real-world traffic scenarios. Videos are labeled with event type (collision/near-collision vs. normal driving), environmental conditions (lighting conditions and weather), and scene type (urban, rural, highway, etc.). For collision and near-collision cases, additional temporal labels are provided, including the precise moment of the event and the alert time, marking when the collision first becomes predictable. To advance research on accident prediction, we introduce the Nexar Dashcam Collision Prediction Challenge, a public competition on top of this dataset. Participants are tasked with developing machine learning models that predict the likelihood of an imminent collision, given an input video. Model performance is evaluated using the average precision (AP) computed across multiple intervals before the accident (i.e. 500 ms, 1000 ms, and 1500 ms prior to the event), emphasizing the importance of early and reliable predictions. The dataset is released under an open license with restrictions on unethical use, ensuring responsible research and innovation.
Authors: Jiangtong Zhu, Zhao Yang, Yinan Shi, Jianwu Fang, Jianru Xue
Abstract: Online vector map construction based on visual data can bypass the processes of data collection, post-processing, and manual annotation required by traditional map construction, which significantly enhances map-building efficiency. However, existing work treats the online mapping task as a local range perception task, overlooking the spatial scalability required for map construction. We propose IC-Mapper, an instance-centric online mapping framework, which comprises two primary components: 1) Instance-centric temporal association module: For the detection queries of adjacent frames, we measure them in both feature and geometric dimensions to obtain the matching correspondence between instances across frames. 2) Instance-centric spatial fusion module: We perform point sampling on the historical global map from a spatial dimension and integrate it with the detection results of instances corresponding to the current frame to achieve real-time expansion and update of the map. Based on the nuScenes dataset, we evaluate our approach on detection, tracking, and global mapping metrics. Experimental results demonstrate the superiority of IC-Mapper against other state-of-the-art methods. Code will be released on https://github.com/Brickzhuantou/IC-Mapper.
Authors: Gal Yona, Roy Velich, Ron Kimmel, Ehud Rivlin
Abstract: Classical shape descriptors such as Heat Kernel Signature (HKS), Wave Kernel Signature (WKS), and Signature of Histograms of OrienTations (SHOT), while widely used in shape analysis, exhibit sensitivity to mesh connectivity, sampling patterns, and topological noise. While differential geometry offers a promising alternative through its theory of differential invariants, which are theoretically guaranteed to be robust shape descriptors, the computation of these invariants on discrete meshes often leads to unstable numerical approximations, limiting their practical utility. We present a self-supervised learning approach for extracting geometric features from 3D surfaces. Our method combines synthetic data generation with a neural architecture designed to learn sampling-invariant features. By integrating our features into existing shape correspondence frameworks, we demonstrate improved performance on standard benchmarks including FAUST, SCAPE, TOPKIDS, and SHREC'16, showing particular robustness to topological noise and partial shapes.
Authors: Devanish N. Kamtam, Joseph B. Shrager, Satya Deepya Malla, Xiaohan Wang, Nicole Lin, Juan J. Cardona, Serena Yeung-Levy, Clarence Hu
Abstract: Background: We evaluate SAM 2 for surgical scene understanding by examining its semantic segmentation capabilities for organs/tissues both in zero-shot scenarios and after fine-tuning. Methods: We utilized five public datasets to evaluate and fine-tune SAM 2 for segmenting anatomical tissues in surgical videos/images. Fine-tuning was applied to the image encoder and mask decoder. We limited training subsets from 50 to 400 samples per class to better model real-world constraints with data acquisition. The impact of dataset size on fine-tuning performance was evaluated with weighted mean Dice coefficient (WMDC), and the results were also compared against previously reported state-of-the-art (SOTA) results. Results: SurgiSAM 2, a fine-tuned SAM 2 model, demonstrated significant improvements in segmentation performance, achieving a 17.9% relative WMDC gain compared to the baseline SAM 2. Increasing prompt points from 1 to 10 and training data scale from 50/class to 400/class enhanced performance; the best WMDC of 0.92 on the validation subset was achieved with 10 prompt points and 400 samples per class. On the test subset, this model outperformed prior SOTA methods in 24/30 (80%) of the classes with a WMDC of 0.91 using 10-point prompts. Notably, SurgiSAM 2 generalized effectively to unseen organ classes, achieving SOTA on 7/9 (77.8%) of them. Conclusion: SAM 2 achieves remarkable zero-shot and fine-tuned performance for surgical scene segmentation, surpassing prior SOTA models across several organ classes of diverse datasets. This suggests immense potential for enabling automated/semi-automated annotation pipelines, thereby decreasing the burden of annotations facilitating several surgical applications.
Authors: Yaopei Zeng, Yuanpu Cao, Lu Lin
Abstract: The growing accessibility of diffusion models has revolutionized image editing but also raised significant concerns about unauthorized modifications, such as misinformation and plagiarism. Existing countermeasures largely rely on adversarial perturbations designed to disrupt diffusion model outputs. However, these approaches are found to be easily neutralized by simple image preprocessing techniques, such as compression and noise addition. To address this limitation, we propose GuardDoor, a novel and robust protection mechanism that fosters collaboration between image owners and model providers. Specifically, the model provider participating in the mechanism fine-tunes the image encoder to embed a protective backdoor, allowing image owners to request the attachment of imperceptible triggers to their images. When unauthorized users attempt to edit these protected images with this diffusion model, the model produces meaningless outputs, reducing the risk of malicious image editing. Our method demonstrates enhanced robustness against image preprocessing operations and is scalable for large-scale deployment. This work underscores the potential of cooperative frameworks between model providers and image owners to safeguard digital content in the era of generative AI.
Authors: Aurelio Noca, Xianmei Lei, Jonathan Becktor, Jeffrey Edlund, Anna Sabel, Patrick Spieler, Curtis Padgett, Alexandre Alahi, Deegan Atha
Abstract: Autonomous off-road navigation faces challenges due to diverse, unstructured environments, requiring robust perception with both geometric and semantic understanding. However, scarce densely labeled semantic data limits generalization across domains. Simulated data helps, but introduces domain adaptation issues. We propose COARSE, a semi-supervised domain adaptation framework for off-road semantic segmentation, leveraging sparse, coarse in-domain labels and densely labeled out-of-domain data. Using pretrained vision transformers, we bridge domain gaps with complementary pixel-level and patch-level decoders, enhanced by a collaborative pseudo-labeling strategy on unlabeled data. Evaluations on RUGD and Rellis-3D datasets show significant improvements of 9.7\% and 8.4\% respectively, versus only using coarse data. Tests on real-world off-road vehicle data in a multi-biome setting further demonstrate COARSE's applicability.
Authors: Wenhui Zhu, Xin Li, Xiwen Chen, Peijie Qiu, Vamsi Krishna Vasa, Xuanzhao Dong, Yanxi Chen, Natasha Lepore, Oana Dumitrascu, Yi Su, Yalin Wang
Abstract: Recently, Multimodal Large Language Models (MLLMs) have gained significant attention for their remarkable ability to process and analyze non-textual data, such as images, videos, and audio. Notably, several adaptations of general-domain MLLMs to the medical field have been explored, including LLaVA-Med. However, these medical adaptations remain insufficiently advanced in understanding and interpreting retinal images. In contrast, medical experts emphasize the importance of quantitative analyses for disease detection and interpretation. This underscores a gap between general-domain and medical-domain MLLMs: while general-domain MLLMs excel in broad applications, they lack the specialized knowledge necessary for precise diagnostic and interpretative tasks in the medical field. To address these challenges, we introduce \textit{RetinalGPT}, a multimodal conversational assistant for clinically preferred quantitative analysis of retinal images. Specifically, we achieve this by compiling a large retinal image dataset, developing a novel data pipeline, and employing customized visual instruction tuning to enhance both retinal analysis and enrich medical knowledge. In particular, RetinalGPT outperforms MLLM in the generic domain by a large margin in the diagnosis of retinal diseases in 8 benchmark retinal datasets. Beyond disease diagnosis, RetinalGPT features quantitative analyses and lesion localization, representing a pioneering step in leveraging LLMs for an interpretable and end-to-end clinical research framework. The code is available at https://github.com/Retinal-Research/RetinalGPT
Authors: Amin Karimi, Charalambos Poullis
Abstract: Few-shot semantic segmentation (FSS) aims to enable models to segment novel/unseen object classes using only a limited number of labeled examples. However, current FSS methods frequently struggle with generalization due to incomplete and biased feature representations, especially when support images do not capture the full appearance variability of the target class. To improve the FSS pipeline, we propose a novel framework that utilizes large language models (LLMs) to adapt general class semantic information to the query image. Furthermore, the framework employs dense pixel-wise matching to identify similarities between query and support images, resulting in enhanced FSS performance. Inspired by reasoning-based segmentation frameworks, our method, named DSV-LFS, introduces an additional token into the LLM vocabulary, allowing a multimodal LLM to generate a "semantic prompt" from class descriptions. In parallel, a dense matching module identifies visual similarities between the query and support images, generating a "visual prompt". These prompts are then jointly employed to guide the prompt-based decoder for accurate segmentation of the query image. Comprehensive experiments on the benchmark datasets Pascal-$5^{i}$ and COCO-$20^{i}$ demonstrate that our framework achieves state-of-the-art performance-by a significant margin-demonstrating superior generalization to novel classes and robustness across diverse scenarios. The source code is available at \href{https://github.com/aminpdik/DSV-LFS}{https://github.com/aminpdik/DSV-LFS}
URLs: https://github.com/aminpdik/DSV-LFS, https://github.com/aminpdik/DSV-LFS
Authors: Kequan Chen, Pan Liu, Yuxuan Wang, David Z. W. Wang, Yifan Dai, Zhibin Li
Abstract: Accurate prediction of traffic crash risks for individual vehicles is essential for enhancing vehicle safety. While significant attention has been given to traffic crash risk prediction, existing studies face two main challenges: First, due to the scarcity of individual vehicle data before crashes, most models rely on hypothetical scenarios deemed dangerous by researchers. This raises doubts about their applicability to actual pre-crash conditions. Second, some crash risk prediction frameworks were learned from dashcam videos. Although such videos capture the pre-crash behavior of individual vehicles, they often lack critical information about the movements of surrounding vehicles. However, the interaction between a vehicle and its surrounding vehicles is highly influential in crash occurrences. To overcome these challenges, we propose a novel non-stationary extreme value theory (EVT), where the covariate function is optimized in a nonlinear fashion using a graph attention network. The EVT component incorporates the stochastic nature of crashes through probability distribution, which enhances model interpretability. Notably, the nonlinear covariate function enables the model to capture the interactive behavior between the target vehicle and its multiple surrounding vehicles, facilitating crash risk prediction across different driving tasks. We train and test our model using 100 sets of vehicle trajectory data before real crashes, collected via drones over three years from merging and weaving segments. We demonstrate that our model successfully learns micro-level precursors of crashes and fits a more accurate distribution with the aid of the nonlinear covariate function. Our experiments on the testing dataset show that the proposed model outperforms existing models by providing more accurate predictions for both rear-end and sideswipe crashes simultaneously.
Authors: Wanglong Lu, Lingming Su, Jingjing Zheng, Vin\'icius Veloso de Melo, Farzaneh Shoeleh, John Hawkin, Terrence Tricco, Hanli Zhao, Xianta Jiang
Abstract: Digital versions of real-world text documents often suffer from issues like environmental corrosion of the original document, low-quality scanning, or human interference. Existing document restoration and inpainting methods typically struggle with generalizing to unseen document styles and handling high-resolution images. To address these challenges, we introduce TextDoctor, a novel unified document image inpainting method. Inspired by human reading behavior, TextDoctor restores fundamental text elements from patches and then applies diffusion models to entire document images instead of training models on specific document types. To handle varying text sizes and avoid out-of-memory issues, common in high-resolution documents, we propose using structure pyramid prediction and patch pyramid diffusion models. These techniques leverage multiscale inputs and pyramid patches to enhance the quality of inpainting both globally and locally. Extensive qualitative and quantitative experiments on seven public datasets validated that TextDoctor outperforms state-of-the-art methods in restoring various types of high-resolution document images.
Authors: Aocheng Li, James R. Zimmer-Dauphinee, Rajesh Kalyanam, Ian Lindsay, Parker VanValkenburgh, Steven Wernke, Daniel Aliaga
Abstract: Point cloud completion helps restore partial incomplete point clouds suffering occlusions. Current self-supervised methods fail to give high fidelity completion for large objects with missing surfaces and unbalanced distribution of available points. In this paper, we present a novel method for restoring large-scale point clouds with limited and imbalanced ground-truth. Using rough boundary annotations for a region of interest, we project the original point clouds into a multiple-center-of-projection (MCOP) image, where fragments are projected to images of 5 channels (RGB, depth, and rotation). Completion of the original point cloud is reduced to inpainting the missing pixels in the MCOP images. Due to lack of complete structures and an unbalanced distribution of existing parts, we develop a self-supervised scheme which learns to infill the MCOP image with points resembling existing "complete" patches. Special losses are applied to further enhance the regularity and consistency of completed MCOP images, which is mapped back to 3D to form final restoration. Extensive experiments demonstrate the superiority of our method in completing 600+ incomplete and unbalanced archaeological structures in Peru.
Authors: Xihan Wang, Dianyi Yang, Yu Gao, Yufeng Yue, Yi Yang, Mengyin Fu
Abstract: Recent advancements in 3D Gaussian Splatting(3DGS) have significantly improved semantic scene understanding, enabling natural language queries to localize objects within a scene. However, existing methods primarily focus on embedding compressed CLIP features to 3D Gaussians, suffering from low object segmentation accuracy and lack spatial reasoning capabilities. To address these limitations, we propose GaussianGraph, a novel framework that enhances 3DGS-based scene understanding by integrating adaptive semantic clustering and scene graph generation. We introduce a "Control-Follow" clustering strategy, which dynamically adapts to scene scale and feature distribution, avoiding feature compression and significantly improving segmentation accuracy. Additionally, we enrich scene representation by integrating object attributes and spatial relations extracted from 2D foundation models. To address inaccuracies in spatial relationships, we propose 3D correction modules that filter implausible relations through spatial consistency verification, ensuring reliable scene graph construction. Extensive experiments on three datasets demonstrate that GaussianGraph outperforms state-of-the-art methods in both semantic segmentation and object grounding tasks, providing a robust solution for complex scene understanding and interaction.
Authors: Zhong Ji, Weilong Cao, Yan Zhang, Yanwei Pang, Jungong Han, Xuelong Li
Abstract: Diffusion models has emerged as a powerful framework for tasks like image controllable generation and dense prediction. However, existing models often struggle to capture underlying semantics (e.g., edges, textures, shapes) and effectively utilize in-context learning, limiting their contextual understanding and image generation quality. Additionally, high computational costs and slow inference speeds hinder their real-time applicability. To address these challenges, we propose Underlying Semantic Diffusion (US-Diffusion), an enhanced diffusion model that boosts underlying semantics learning, computational efficiency, and in-context learning capabilities on multi-task scenarios. We introduce Separate & Gather Adapter (SGA), which decouples input conditions for different tasks while sharing the architecture, enabling better in-context learning and generalization across diverse visual domains. We also present a Feedback-Aided Learning (FAL) framework, which leverages feedback signals to guide the model in capturing semantic details and dynamically adapting to task-specific contextual cues. Furthermore, we propose a plug-and-play Efficient Sampling Strategy (ESS) for dense sampling at time steps with high-noise levels, which aims at optimizing training and inference efficiency while maintaining strong in-context learning performance. Experimental results demonstrate that US-Diffusion outperforms the state-of-the-art method, achieving an average reduction of 7.47 in FID on Map2Image tasks and an average reduction of 0.026 in RMSE on Image2Map tasks, while achieving approximately 9.45 times faster inference speed. Our method also demonstrates superior training efficiency and in-context learning capabilities, excelling in new datasets and tasks, highlighting its robustness and adaptability across diverse visual domains.
Authors: Haiyang Yu, Jinghui Lu, Yanjie Wang, Yang Li, Han Wang, Can Huang, Bin Li
Abstract: The advent of Large Vision-Language Models (LVLMs) has advanced the video-based tasks, such as video captioning and video understanding. Some previous research indicates that taking texts in videos as input can further improve the performance of video understanding. As a type of indispensable information in short videos or movies, subtitles can assist LVLMs to better understand videos. Most existing methods for video subtitle extraction are based on a multi-stage framework, handling each frame independently. They can hardly exploit the temporal information of videos. Although some LVLMs exhibit the robust OCR capability, predicting accurate timestamps for subtitle texts is still challenging. In this paper, we propose an End-to-end Video Subtitle Extraction method, called EVE, which consists of three modules: a vision encoder, an adapter module, and a large language model. To effectively compress the visual tokens from the vision encoder, we propose a novel adapter InterleavedVT to interleave two modalities. It contains a visual compressor and a textual region compressor. The proposed InterleavedVT exploits both the merits of average pooling and Q-Former in token compression. Taking the temporal information of videos into account, we introduce a sliding-window mechanism in the textual region compressor. To benchmark the video subtitle extraction task, we propose a large dataset ViSa including 2.5M videos. Extensive experiments on ViSa demonstrate that the proposed EVE can outperform existing open-sourced tools and LVLMs.
Authors: Yunxiao Shi, Hong Cai, Amin Ansari, Fatih Porikli
Abstract: 3D occupancy prediction has recently emerged as a new paradigm for holistic 3D scene understanding and provides valuable information for downstream planning in autonomous driving. Most existing methods, however, are computationally expensive, requiring costly attention-based 2D-3D transformation and 3D feature processing. In this paper, we present a novel 3D occupancy prediction approach, H3O, which features highly efficient architecture designs that incur a significantly lower computational cost as compared to the current state-of-the-art methods. In addition, to compensate for the ambiguity in ground-truth 3D occupancy labels, we advocate leveraging auxiliary tasks to complement the direct 3D supervision. In particular, we integrate multi-camera depth estimation, semantic segmentation, and surface normal estimation via differentiable volume rendering, supervised by corresponding 2D labels that introduces rich and heterogeneous supervision signals. We conduct extensive experiments on the Occ3D-nuScenes and SemanticKITTI benchmarks that demonstrate the superiority of our proposed H3O.
Authors: Feng Ni, Kui Huang, Yao Lu, Wenyu Lv, Guanzhong Wang, Zeyu Chen, Yi Liu
Abstract: With the rapid advancement of digitalization, various document images are being applied more extensively in production and daily life, and there is an increasingly urgent need for fast and accurate parsing of the content in document images. Therefore, this report presents PP-DocBee, a novel multimodal large language model designed for end-to-end document image understanding. First, we develop a data synthesis strategy tailored to document scenarios in which we build a diverse dataset to improve the model generalization. Then, we apply a few training techniques, including dynamic proportional sampling, data preprocessing, and OCR postprocessing strategies. Extensive evaluations demonstrate the superior performance of PP-DocBee, achieving state-of-the-art results on English document understanding benchmarks and even outperforming existing open source and commercial models in Chinese document understanding. The source code and pre-trained models are publicly available at \href{https://github.com/PaddlePaddle/PaddleMIX}{https://github.com/PaddlePaddle/PaddleMIX}.
URLs: https://github.com/PaddlePaddle/PaddleMIX, https://github.com/PaddlePaddle/PaddleMIX
Authors: Ziqi Ni, Ao Fu, Yi Zhou
Abstract: Achieving high-fidelity lip-speech synchronization in audio-driven talking portrait synthesis remains challenging. While multi-stage pipelines or diffusion models yield high-quality results, they suffer from high computational costs. Some approaches perform well on specific individuals with low resources, yet still exhibit mismatched lip movements. The aforementioned methods are modeled in the pixel domain. We observed that there are noticeable discrepancies in the frequency domain between the synthesized talking videos and natural videos. Currently, no research on talking portrait synthesis has considered this aspect. To address this, we propose a FREquency-modulated, high-fidelity, and real-time Audio-driven talKing portrait synthesis framework, named FREAK, which models talking portraits from the frequency domain perspective, enhancing the fidelity and naturalness of the synthesized portraits. FREAK introduces two novel frequency-based modules: 1) the Visual Encoding Frequency Modulator (VEFM) to couple multi-scale visual features in the frequency domain, better preserving visual frequency information and reducing the gap in the frequency spectrum between synthesized and natural frames. and 2) the Audio Visual Frequency Modulator (AVFM) to help the model learn the talking pattern in the frequency domain and improve audio-visual synchronization. Additionally, we optimize the model in both pixel domain and frequency domain jointly. Furthermore, FREAK supports seamless switching between one-shot and video dubbing settings, offering enhanced flexibility. Due to its superior performance, it can simultaneously support high-resolution video results and real-time inference. Extensive experiments demonstrate that our method synthesizes high-fidelity talking portraits with detailed facial textures and precise lip synchronization in real-time, outperforming state-of-the-art methods.
Authors: Qing Chang, Wei Dai, Zhihao Shuai, Limin Yu, Yutao Yue
Abstract: Naturalistic driving action recognition is essential for vehicle cabin monitoring systems. However, the complexity of real-world backgrounds presents significant challenges for this task, and previous approaches have struggled with practical implementation due to their limited ability to observe subtle behavioral differences and effectively learn inter-frame features from video. In this paper, we propose a novel Spatial-Temporal Perception (STP) architecture that emphasizes both temporal information and spatial relationships between key objects, incorporating a causal decoder to perform behavior recognition and temporal action localization. Without requiring multimodal input, STP directly extracts temporal and spatial distance features from RGB video clips. Subsequently, these dual features are jointly encoded by maximizing the expected likelihood across all possible permutations of the factorization order. By integrating temporal and spatial features at different scales, STP can perceive subtle behavioral changes in challenging scenarios. Additionally, we introduce a causal-aware module to explore relationships between video frame features, significantly enhancing detection efficiency and performance. We validate the effectiveness of our approach using two publicly available driver distraction detection benchmarks. The results demonstrate that our framework achieves state-of-the-art performance.
Authors: Idris O. Sunmola, Zhenjun Zhao, Samuel Schmidgall, Yumeng Wang, Paul Maria Scheikl, Axel Krieger
Abstract: Accurate geometric reconstruction of deformable tissues in monocular endoscopic video remains a fundamental challenge in robot-assisted minimally invasive surgery. Although recent volumetric and point primitive methods based on neural radiance fields (NeRF) and 3D Gaussian primitives have efficiently rendered surgical scenes, they still struggle with handling artifact-free tool occlusions and preserving fine anatomical details. These limitations stem from unrestricted Gaussian scaling and insufficient surface alignment constraints during reconstruction. To address these issues, we introduce Surgical Gaussian Surfels (SGS), which transforms anisotropic point primitives into surface-aligned elliptical splats by constraining the scale component of the Gaussian covariance matrix along the view-aligned axis. We predict accurate surfel motion fields using a lightweight Multi-Layer Perceptron (MLP) coupled with locality constraints to handle complex tissue deformations. We use homodirectional view-space positional gradients to capture fine image details by splitting Gaussian Surfels in over-reconstructed regions. In addition, we define surface normals as the direction of the steepest density change within each Gaussian surfel primitive, enabling accurate normal estimation without requiring monocular normal priors. We evaluate our method on two in-vivo surgical datasets, where it outperforms current state-of-the-art methods in surface geometry, normal map quality, and rendering efficiency, while remaining competitive in real-time rendering performance. We make our code available at https://github.com/aloma85/SurgicalGaussianSurfels
Authors: Shuojue Yang, Zijian Wu, Mingxuan Hong, Qian Li, Daiyun Shen, Septimiu E. Salcudean, Yueming Jin
Abstract: Real2Sim is becoming increasingly important with the rapid development of surgical artificial intelligence (AI) and autonomy. In this work, we propose a novel Real2Sim methodology, \textit{Instrument-Splatting}, that leverages 3D Gaussian Splatting to provide fully controllable 3D reconstruction of surgical instruments from monocular surgical videos. To maintain both high visual fidelity and manipulability, we introduce a geometry pre-training to bind Gaussian point clouds on part mesh with accurate geometric priors and define a forward kinematics to control the Gaussians as flexible as real instruments. Afterward, to handle unposed videos, we design a novel instrument pose tracking method leveraging semantics-embedded Gaussians to robustly refine per-frame instrument poses and joint states in a render-and-compare manner, which allows our instrument Gaussian to accurately learn textures and reach photorealistic rendering. We validated our method on 2 publicly released surgical videos and 4 videos collected on ex vivo tissues and green screens. Quantitative and qualitative evaluations demonstrate the effectiveness and superiority of the proposed method.
Authors: Sheikh Moonwara Anjum Monisha, Ratun Rahman
Abstract: One of the primary challenges in medical diagnostics is the accurate and efficient use of magnetic resonance imaging (MRI) for the detection of brain tumors. But the current machine learning (ML) approaches have two major limitations, data privacy and high latency. To solve the problem, in this work we propose a federated learning architecture for a better accurate brain tumor detection incorporating the YOLOv11 algorithm. In contrast to earlier methods of centralized learning, our federated learning approach protects the underlying medical data while supporting cooperative deep learning model training across multiple institutions. To allow the YOLOv11 model to locate and identify tumor areas, we adjust it to handle MRI data. To ensure robustness and generalizability, the model is trained and tested on a wide range of MRI data collected from several anonymous medical facilities. The results indicate that our method significantly maintains higher accuracy than conventional approaches.
Authors: Haoran Wang, Lian Huai, Wenbin Li, Lei Qi, Xingqun Jiang, Yinghuan Shi
Abstract: We have witnessed remarkable progress in foundation models in vision tasks. Currently, several recent works have utilized the segmenting anything model (SAM) to boost the segmentation performance in medical images, where most of them focus on training an adaptor for fine-tuning a large amount of pixel-wise annotated medical images following a fully supervised manner. In this paper, to reduce the labeling cost, we investigate a novel weakly-supervised SAM-based segmentation model, namely WeakMedSAM. Specifically, our proposed WeakMedSAM contains two modules: 1) to mitigate severe co-occurrence in medical images, a sub-class exploration module is introduced to learn accurate feature representations. 2) to improve the quality of the class activation maps, our prompt affinity mining module utilizes the prompt capability of SAM to obtain an affinity map for random-walk refinement. Our method can be applied to any SAM-like backbone, and we conduct experiments with SAMUS and EfficientSAM. The experimental results on three popularly-used benchmark datasets, i.e., BraTS 2019, AbdomenCT-1K, and MSD Cardiac dataset, show the promising results of our proposed WeakMedSAM. Our code is available at https://github.com/wanghr64/WeakMedSAM.
Authors: Masoumeh Zareapoor, Pourya Shamsolmoali, Huiyu Zhou, Yue Lu, Salvador Garc\'ia
Abstract: The Detection Transformer (DETR), by incorporating the Hungarian algorithm, has significantly simplified the matching process in object detection tasks. This algorithm facilitates optimal one-to-one matching of predicted bounding boxes to ground-truth annotations during training. While effective, this strict matching process does not inherently account for the varying densities and distributions of objects, leading to suboptimal correspondences such as failing to handle multiple detections of the same object or missing small objects. To address this, we propose the Regularized Transport Plan (RTP). RTP introduces a flexible matching strategy that captures the cost of aligning predictions with ground truths to find the most accurate correspondences between these sets. By utilizing the differentiable Sinkhorn algorithm, RTP allows for soft, fractional matching rather than strict one-to-one assignments. This approach enhances the model's capability to manage varying object densities and distributions effectively. Our extensive evaluations on the MS-COCO and VOC benchmarks demonstrate the effectiveness of our approach. RTP-DETR, surpassing the performance of the Deform-DETR and the recently introduced DINO-DETR, achieving absolute gains in mAP of +3.8% and +1.7%, respectively.
Authors: Chunnan Shang, Zhizhong Wang, Hongwei Wang, Xiangming Meng
Abstract: Attention-based arbitrary style transfer methods, including CNN-based, Transformer-based, and Diffusion-based, have flourished and produced high-quality stylized images. However, they perform poorly on the content and style images with the same semantics, i.e., the style of the corresponding semantic region of the generated stylized image is inconsistent with that of the style image. We argue that the root cause lies in their failure to consider the relationship between local regions and semantic regions. To address this issue, we propose a plug-and-play semantic continuous-sparse attention, dubbed SCSA, for arbitrary semantic style transfer -- each query point considers certain key points in the corresponding semantic region. Specifically, semantic continuous attention ensures each query point fully attends to all the continuous key points in the same semantic region that reflect the overall style characteristics of that region; Semantic sparse attention allows each query point to focus on the most similar sparse key point in the same semantic region that exhibits the specific stylistic texture of that region. By combining the two modules, the resulting SCSA aligns the overall style of the corresponding semantic regions while transferring the vivid textures of these regions. Qualitative and quantitative results prove that SCSA enables attention-based arbitrary style transfer methods to produce high-quality semantic stylized images.
Authors: Alan Luo, Kaiwen Yuan
Abstract: Vision Transformers (ViTs) have demonstrated exceptional performance in various vision tasks. However, they tend to underperform on smaller datasets due to their inherent lack of inductive biases. Current approaches address this limitation implicitly-often by pairing ViTs with pretext tasks or by distilling knowledge from convolutional neural networks (CNNs) to strengthen the prior. In contrast, Self-Organizing Maps (SOMs), a widely adopted self-supervised framework, are inherently structured to preserve topology and spatial organization, making them a promising candidate to directly address the limitations of ViTs in limited or small training datasets. Despite this potential, equipping SOMs with modern deep learning architectures remains largely unexplored. In this study, we conduct a novel exploration on how Vision Transformers (ViTs) and Self-Organizing Maps (SOMs) can empower each other, aiming to bridge this critical research gap. Our findings demonstrate that these architectures can synergistically enhance each other, leading to significantly improved performance in both unsupervised and supervised tasks. Code will be publicly available.
Authors: Qianliang Wu, Haobo Jiang, Yaqing Ding, Lei Luo, Jin Xie, Jian Yang
Abstract: Establishing reliable correspondences is crucial for all registration tasks, including 2D image registration, 3D point cloud registration, and 2D-3D image-to-point cloud registration. However, these tasks are often complicated by challenges such as scale inconsistencies, symmetry, and large deformations, which can lead to ambiguous matches. Previous feature-based and correspondence-based methods typically rely on geometric or semantic features to generate or polish initial potential correspondences. Some methods typically leverage specific geometric priors, such as topological preservation, to devise diverse and innovative strategies tailored to a given enhancement goal, which cannot be exhaustively enumerated. Additionally, many previous approaches rely on a single-step prediction head, which can struggle with local minima in complex matching scenarios. To address these challenges, we introduce an innovative paradigm that leverages a diffusion model in matrix space for robust matching matrix estimation. Our model treats correspondence estimation as a denoising diffusion process in the matching matrix space, gradually refining the intermediate matching matrix to the optimal one. Specifically, we apply the diffusion model in the doubly stochastic matrix space for 3D-3D and 2D-3D registration tasks. In the 2D image registration task, we deploy the diffusion model in a matrix subspace where dual-softmax projection regularization is applied. For all three registration tasks, we provide adaptive matching matrix embedding implementations tailored to the specific characteristics of each task while maintaining a consistent "match-to-warp" encoding pattern. Furthermore, we adopt a lightweight design for the denoising module. In inference, once points or image features are extracted and fixed, this module performs multi-step denoising predictions through reverse sampling.
Authors: Jindong Jiang, Xiuyu Li, Zhijian Liu, Muyang Li, Guo Chen, Zhiqi Li, De-An Huang, Guilin Liu, Zhiding Yu, Kurt Keutzer, Sungjin Ahn, Jan Kautz, Hongxu Yin, Yao Lu, Song Han, Wonmin Byeon
Abstract: Recent advances in video-based multimodal large language models (Video-LLMs) have significantly improved video understanding by processing videos as sequences of image frames. However, many existing methods treat frames independently in the vision backbone, lacking explicit temporal modeling, which limits their ability to capture dynamic patterns and efficiently handle long videos. To address these limitations, we introduce STORM (\textbf{S}patiotemporal \textbf{TO}ken \textbf{R}eduction for \textbf{M}ultimodal LLMs), a novel architecture incorporating a dedicated temporal encoder between the image encoder and the LLM. Our temporal encoder leverages the Mamba State Space Model to integrate temporal information into image tokens, generating enriched representations that preserve inter-frame dynamics across the entire video sequence. This enriched encoding not only enhances video reasoning capabilities but also enables effective token reduction strategies, including test-time sampling and training-based temporal and spatial pooling, substantially reducing computational demands on the LLM without sacrificing key temporal information. By integrating these techniques, our approach simultaneously reduces training and inference latency while improving performance, enabling efficient and robust video understanding over extended temporal contexts. Extensive evaluations show that STORM achieves state-of-the-art results across various long video understanding benchmarks (more than 5\% improvement on MLVU and LongVideoBench) while reducing the computation costs by up to $8\times$ and the decoding latency by 2.4-2.9$\times$ for the fixed numbers of input frames. Project page is available at https://research.nvidia.com/labs/lpr/storm
Authors: Jie Liu, Tiexin Qin, Hui Liu, Yilei Shi, Lichao Mou, Xiao Xiang Zhu, Shiqi Wang, Haoliang Li
Abstract: In this work, we address the challenge of adaptive pediatric Left Ventricular Ejection Fraction (LVEF) assessment. While Test-time Training (TTT) approaches show promise for this task, they suffer from two significant limitations. Existing TTT works are primarily designed for classification tasks rather than continuous value regression, and they lack mechanisms to handle the quasi-periodic nature of cardiac signals. To tackle these issues, we propose a novel \textbf{Q}uasi-\textbf{P}eriodic \textbf{A}daptive \textbf{R}egression with \textbf{T}est-time Training (Q-PART) framework. In the training stage, the proposed Quasi-Period Network decomposes the echocardiogram into periodic and aperiodic components within latent space by combining parameterized helix trajectories with Neural Controlled Differential Equations. During inference, our framework further employs a variance minimization strategy across image augmentations that simulate common quality issues in echocardiogram acquisition, along with differential adaptation rates for periodic and aperiodic components. Theoretical analysis is provided to demonstrate that our variance minimization objective effectively bounds the regression error under mild conditions. Furthermore, extensive experiments across three pediatric age groups demonstrate that Q-PART not only significantly outperforms existing approaches in pediatric LVEF prediction, but also exhibits strong clinical screening capability with high mAUROC scores (up to 0.9747) and maintains gender-fair performance across all metrics, validating its robustness and practical utility in pediatric echocardiography analysis.
Authors: Junchi Feng, Giles Hamilton-Fletcher, Nikhil Ballem, Michael Batavia, Yifei Wang, Jiuling Zhong, Maurizio Porfiri, John-Ross Rizzo
Abstract: Navigating urban environments poses significant challenges for people with disabilities, particularly those with blindness and low vision. Environments with dynamic and unpredictable elements like construction sites are especially challenging. Construction sites introduce hazards like uneven surfaces, obstructive barriers, hazardous materials, and excessive noise, and they can alter routing, complicating safe mobility. Existing assistive technologies are limited, as navigation apps do not account for construction sites during trip planning, and detection tools that attempt hazard recognition struggle to address the extreme variability of construction paraphernalia. This study introduces a novel computer vision-based system that integrates open-vocabulary object detection, a YOLO-based scaffolding-pole detection model, and an optical character recognition (OCR) module to comprehensively identify and interpret construction site elements for assistive navigation. In static testing across seven construction sites, the system achieved an overall accuracy of 88.56\%, reliably detecting objects from 2m to 10m within a 0$^\circ$ -- 75$^\circ$ angular offset. At closer distances (2--4m), the detection rate was 100\% at all tested angles. At
Authors: Yating Liu, Zimo Liu, Xiangyuan Lan, Wenming Yang, Yaowei Li, Qingmin Liao
Abstract: Text-based person retrieval (TPR) has gained significant attention as a fine-grained and challenging task that closely aligns with practical applications. Tailoring CLIP to person domain is now a emerging research topic due to the abundant knowledge of vision-language pretraining, but challenges still remain during fine-tuning: (i) Previous full-model fine-tuning in TPR is computationally expensive and prone to overfitting.(ii) Existing parameter-efficient transfer learning (PETL) for TPR lacks of fine-grained feature extraction. To address these issues, we propose Domain-Aware Mixture-of-Adapters (DM-Adapter), which unifies Mixture-of-Experts (MOE) and PETL to enhance fine-grained feature representations while maintaining efficiency. Specifically, Sparse Mixture-of-Adapters is designed in parallel to MLP layers in both vision and language branches, where different experts specialize in distinct aspects of person knowledge to handle features more finely. To promote the router to exploit domain information effectively and alleviate the routing imbalance, Domain-Aware Router is then developed by building a novel gating function and injecting learnable domain-aware prompts. Extensive experiments show that our DM-Adapter achieves state-of-the-art performance, outperforming previous methods by a significant margin.
Authors: Jie Xu, Na Zhao, Gang Niu, Masashi Sugiyama, Xiaofeng Zhu
Abstract: Recently, multi-view learning (MVL) has garnered significant attention due to its ability to fuse discriminative information from multiple views. However, real-world multi-view datasets are often heterogeneous and imperfect, which usually makes MVL methods designed for specific combinations of views lack application potential and limits their effectiveness. To address this issue, we propose a novel robust MVL method (namely RML) with simultaneous representation fusion and alignment. Specifically, we introduce a simple yet effective multi-view transformer fusion network where we transform heterogeneous multi-view data into homogeneous word embeddings, and then integrate multiple views by the sample-level attention mechanism to obtain a fused representation. Furthermore, we propose a simulated perturbation based multi-view contrastive learning framework that dynamically generates the noise and unusable perturbations for simulating imperfect data conditions. The simulated noisy and unusable data obtain two distinct fused representations, and we utilize contrastive learning to align them for learning discriminative and robust representations. Our RML is self-supervised and can also be applied for downstream tasks as a regularization. In experiments, we employ it in unsupervised multi-view clustering, noise-label classification, and as a plug-and-play module for cross-modal hashing retrieval. Extensive comparison experiments and ablation studies validate the effectiveness of RML.
Authors: Chupeng Liu, Runkai Zhao, Weidong Cai
Abstract: Weakly supervised monocular 3D detection, while less annotation-intensive, often struggles to capture the global context required for reliable 3D reasoning. Conventional label-efficient methods focus on object-centric features, neglecting contextual semantic relationships that are critical in complex scenes. In this work, we propose a Context-Aware Weak Supervision for Monocular 3D object detection, namely CA-W3D, to address this limitation in a two-stage training paradigm. Specifically, we first introduce a pre-training stage employing Region-wise Object Contrastive Matching (ROCM), which aligns regional object embeddings derived from a trainable monocular 3D encoder and a frozen open-vocabulary 2D visual grounding model. This alignment encourages the monocular encoder to discriminate scene-specific attributes and acquire richer contextual knowledge. In the second stage, we incorporate a pseudo-label training process with a Dual-to-One Distillation (D2OD) mechanism, which effectively transfers contextual priors into the monocular encoder while preserving spatial fidelity and maintaining computational efficiency during inference. Extensive experiments conducted on the public KITTI benchmark demonstrate the effectiveness of our approach, surpassing the SoTA method over all metrics, highlighting the importance of contextual-aware knowledge in weakly-supervised monocular 3D detection.
Authors: Bodong Zhang, Hamid Manoochehri, Beatrice S. Knudsen, Tolga Tasdizen
Abstract: Weakly supervised multiple instance learning (MIL) is a challenging task given that only bag-level labels are provided, while each bag typically contains multiple instances. This topic has been extensively studied in histopathological image analysis, where labels are usually available only at the whole slide image (WSI) level, while each whole slide image can be divided into thousands of small image patches for training. The dominant MIL approaches take fixed patch features as inputs to address computational constraints and ensure model stability. These features are commonly generated by encoders pre-trained on ImageNet, foundation encoders pre-trained on large datasets, or through self-supervised learning on local datasets. While the self-supervised encoder pre-training on the same dataset as downstream MIL tasks helps mitigate domain shift and generate better features, the bag-level labels are not utilized during the process, and the features of patches from different categories may cluster together, reducing classification performance on MIL tasks. Recently, pre-training with supervised contrastive learning (SupCon) has demonstrated superior performance compared to self-supervised contrastive learning and even end-to-end training on traditional image classification tasks. In this paper, we propose a novel encoder pre-training method for downstream MIL tasks called Weakly Supervised Contrastive Learning (WeakSupCon) that utilizes bag-level labels. In our method, we employ multi-task learning and define distinct contrastive learning losses for samples with different bag labels. Our experiments demonstrate that the features generated using WeakSupCon significantly enhance MIL classification performance compared to self-supervised approaches across three datasets.
Authors: Yufang Liu, Yao Du, Tao Ji, Jianing Wang, Yang Liu, Yuanbin Wu, Aimin Zhou, Mengdi Zhang, Xunliang Cai
Abstract: Recent research has increasingly focused on multimodal mathematical reasoning, particularly emphasizing the creation of relevant datasets and benchmarks. Despite this, the role of visual information in reasoning has been underexplored. Our findings show that existing multimodal mathematical models minimally leverage visual information, and model performance remains largely unaffected by changes to or removal of images in the dataset. We attribute this to the dominance of textual information and answer options that inadvertently guide the model to correct answers. To improve evaluation methods, we introduce the HC-M3D dataset, specifically designed to require image reliance for problem-solving and to challenge models with similar, yet distinct, images that change the correct answer. In testing leading models, their failure to detect these subtle visual differences suggests limitations in current visual perception capabilities. Additionally, we observe that the common approach of improving general VQA capabilities by combining various types of image encoders does not contribute to math reasoning performance. This finding also presents a challenge to enhancing visual reliance during math reasoning. Our benchmark and code would be available at \href{https://github.com/Yufang-Liu/visual_modality_role}{https://github.com/Yufang-Liu/visual\_modality\_role}.
URLs: https://github.com/Yufang-Liu/visual_modality_role, https://github.com/Yufang-Liu/visual\_modality\_role
Authors: Zhiqiang Yan, Zhengxue Wang, Haoye Dong, Jun Li, Jian Yang, Gim Hee Lee
Abstract: We introduce DuCos, a novel depth super-resolution framework grounded in Lagrangian duality theory, offering a flexible integration of multiple constraints and reconstruction objectives to enhance accuracy and robustness. Our DuCos is the first to significantly improve generalization across diverse scenarios with foundation models as prompts. The prompt design consists of two key components: Correlative Fusion (CF) and Gradient Regulation (GR). CF facilitates precise geometric alignment and effective fusion between prompt and depth features, while GR refines depth predictions by enforcing consistency with sharp-edged depth maps derived from foundation models. Crucially, these prompts are seamlessly embedded into the Lagrangian constraint term, forming a synergistic and principled framework. Extensive experiments demonstrate that DuCos outperforms existing state-of-the-art methods, achieving superior accuracy, robustness, and generalization. The source codes and pre-trained models will be publicly available.
Authors: Sara Sangalli, Gary Sarwin, Ertunc Erdil, Carlo Serra, Ender Konukoglu
Abstract: Forecasting surgical instrument trajectories and predicting the next surgical action recently started to attract attention from the research community. Both these tasks are crucial for automation and assistance in endoscopy surgery. Given the safety-critical nature of these tasks, reliable uncertainty quantification is essential. Conformal prediction is a fast-growing and widely recognized framework for uncertainty estimation in machine learning and computer vision, offering distribution-free, theoretically valid prediction intervals. In this work, we explore the application of standard conformal prediction and conformalized quantile regression to estimate uncertainty in forecasting surgical instrument motion, i.e., predicting direction and magnitude of surgical instruments' future motion. We analyze and compare their coverage and interval sizes, assessing the impact of multiple hypothesis testing and correction methods. Additionally, we show how these techniques can be employed to produce useful uncertainty heatmaps. To the best of our knowledge, this is the first study applying conformal prediction to surgical guidance, marking an initial step toward constructing principled prediction intervals with formal coverage guarantees in this domain.
Authors: Fuyang Liu, Shun Lu, Jilin Mei, Yu Hu
Abstract: RGB-Thermal fusion is a potential solution for various weather and light conditions in challenging scenarios. However, plenty of studies focus on designing complex modules to fuse different modalities. With the widespread application of large language models (LLMs), valuable information can be more effectively extracted from natural language. Therefore, we aim to leverage the advantages of large language models to design a structurally simple and highly adaptable multimodal fusion model architecture. We proposed MultimodAl Segmentation with TExt PRompts (MASTER) architecture, which integrates LLM into the fusion of RGB-Thermal multimodal data and allows complex query text to participate in the fusion process. Our model utilizes a dual-path structure to extract information from different modalities of images. Additionally, we employ LLM as the core module for multimodal fusion, enabling the model to generate learnable codebook tokens from RGB, thermal images, and textual information. A lightweight image decoder is used to obtain semantic segmentation results. The proposed MASTER performs exceptionally well in benchmark tests across various automated driving scenarios, yielding promising results.
Authors: Zhanhong Jiang, Md Zahid Hasan, Aditya Balu, Joshua R. Waite, Genyi Huang, Soumik Sarkar
Abstract: Stochastic optimization methods have actively been playing a critical role in modern machine learning algorithms to deliver decent performance. While numerous works have proposed and developed diverse approaches, first-order and second-order methods are in entirely different situations. The former is significantly pivotal and dominating in emerging deep learning but only leads convergence to a stationary point. However, second-order methods are less popular due to their computational intensity in large-dimensional problems. This paper presents a novel method that leverages both the first-order and second-order methods in a unified algorithmic framework, termed FUSE, from which a practical version (PV) is derived accordingly. FUSE-PV stands as a simple yet efficient optimization method involving a switch-over between first and second orders. Additionally, we develop different criteria that determine when to switch. FUSE-PV has provably shown a smaller computational complexity than SGD and Adam. To validate our proposed scheme, we present an ablation study on several simple test functions and show a comparison with baselines for benchmark datasets.
Authors: Xingcan Hu, Wei Wang, Li Xiao
Abstract: In MRI-based mental disorder diagnosis, most previous studies focus on functional connectivity network (FCN) derived from functional MRI (fMRI). However, the small size of annotated fMRI datasets restricts its wide application. Meanwhile, structural MRIs (sMRIs), such as 3D T1-weighted (T1w) MRI, which are commonly used and readily accessible in clinical settings, are often overlooked. To integrate the complementary information from both function and structure for improved diagnostic accuracy, we propose CINP (Contrastive Image-Network Pre-training), a framework that employs contrastive learning between sMRI and FCN. During pre-training, we incorporate masked image modeling and network-image matching to enhance visual representation learning and modality alignment. Since the CINP facilitates knowledge transfer from FCN to sMRI, we introduce network prompting. It utilizes only sMRI from suspected patients and a small amount of FCNs from different patient classes for diagnosing mental disorders, which is practical in real-world clinical scenario. The competitive performance on three mental disorder diagnosis tasks demonstrate the effectiveness of the CINP in integrating multimodal MRI information, as well as the potential of incorporating sMRI into clinical diagnosis using network prompting.
Authors: Haitao Wu, Qing Li, Changqing Zhang, Zhen He, Xiaomin Ying
Abstract: Can our brain signals faithfully reflect the original visual stimuli, even including high-frequency details? Although human perceptual and cognitive capacities enable us to process and remember visual information, these abilities are constrained by several factors, such as limited attentional resources and the finite capacity of visual memory. When visual stimuli are processed by human visual system into brain signals, some information is inevitably lost, leading to a discrepancy known as the \textbf{System GAP}. Additionally, perceptual and cognitive dynamics, along with technical noise in signal acquisition, degrade the fidelity of brain signals relative to the visual stimuli, known as the \textbf{Random GAP}. When encoded brain representations are directly aligned with the corresponding pretrained image features, the System GAP and Random GAP between paired data challenge the model, requiring it to bridge these gaps. However, in the context of limited paired data, these gaps are difficult for the model to learn, leading to overfitting and poor generalization to new data. To address these GAPs, we propose a simple yet effective approach called the \textbf{Uncertainty-aware Blur Prior (UBP)}. It estimates the uncertainty within the paired data, reflecting the mismatch between brain signals and visual stimuli. Based on this uncertainty, UBP dynamically blurs the high-frequency details of the original images, reducing the impact of the mismatch and improving alignment. Our method achieves a top-1 accuracy of \textbf{50.9\%} and a top-5 accuracy of \textbf{79.7\%} on the zero-shot brain-to-image retrieval task, surpassing previous state-of-the-art methods by margins of \textbf{13.7\%} and \textbf{9.8\%}, respectively. Code is available at \href{https://github.com/HaitaoWuTJU/Uncertainty-aware-Blur-Prior}{GitHub}.
URLs: https://github.com/HaitaoWuTJU/Uncertainty-aware-Blur-Prior
Authors: Rui Jiang, Xinghe Fu, Guangcong Zheng, Teng Li, Taiping Yao, Xi Li
Abstract: The rapid advancement of pretrained text-driven diffusion models has significantly enriched applications in image generation and editing. However, as the demand for personalized content editing increases, new challenges emerge especially when dealing with arbitrary objects and complex scenes. Existing methods usually mistakes mask as the object shape prior, which struggle to achieve a seamless integration result. The mostly used inversion noise initialization also hinders the identity consistency towards the target object. To address these challenges, we propose a novel training-free framework that formulates personalized content editing as the optimization of edited images in the latent space, using diffusion models as the energy function guidance conditioned by reference text-image pairs. A coarse-to-fine strategy is proposed that employs text energy guidance at the early stage to achieve a natural transition toward the target class and uses point-to-point feature-level image energy guidance to perform fine-grained appearance alignment with the target object. Additionally, we introduce the latent space content composition to enhance overall identity consistency with the target. Extensive experiments demonstrate that our method excels in object replacement even with a large domain gap, highlighting its potential for high-quality, personalized image editing.
Authors: Yi Xiao, Qiangqiang Yuan, Kui Jiang, Qiang Zhang, Tingting Zheng, Chia-Wen Lin, Liangpei Zhang
Abstract: Spiking neural networks (SNNs) are emerging as a promising alternative to traditional artificial neural networks (ANNs), offering biological plausibility and energy efficiency. Despite these merits, SNNs are frequently hampered by limited capacity and insufficient representation power, yet remain underexplored in remote sensing super-resolution (SR) tasks. In this paper, we first observe that spiking signals exhibit drastic intensity variations across diverse textures, highlighting an active learning state of the neurons. This observation motivates us to apply SNNs for efficient SR of RSIs. Inspired by the success of attention mechanisms in representing salient information, we devise the spiking attention block (SAB), a concise yet effective component that optimizes membrane potentials through inferred attention weights, which, in turn, regulates spiking activity for superior feature representation. Our key contributions include: 1) we bridge the independent modulation between temporal and channel dimensions, facilitating joint feature correlation learning, and 2) we access the global self-similar patterns in large-scale remote sensing imagery to infer spatial attention weights, incorporating effective priors for realistic and faithful reconstruction. Building upon SAB, we proposed SpikeSR, which achieves state-of-the-art performance across various remote sensing benchmarks such as AID, DOTA, and DIOR, while maintaining high computational efficiency. The code of SpikeSR will be available upon paper acceptance.
Authors: Bin Wu, Wuxuan Shi, Jinqiao Wang, Mang Ye
Abstract: Pre-trained Vision-Language Models (VLMs) require Continual Learning (CL) to efficiently update their knowledge and adapt to various downstream tasks without retraining from scratch. However, for VLMs, in addition to the loss of knowledge previously learned from downstream tasks, pre-training knowledge is also corrupted during continual fine-tuning. This issue is exacerbated by the unavailability of original pre-training data, leaving VLM's generalization ability degrading. In this paper, we propose GIFT, a novel continual fine-tuning approach that utilizes synthetic data to overcome catastrophic forgetting in VLMs. Taking advantage of recent advances in text-to-image synthesis, we employ a pre-trained diffusion model to recreate both pre-training and learned downstream task data. In this way, the VLM can revisit previous knowledge through distillation on matching diffusion-generated images and corresponding text prompts. Leveraging the broad distribution and high alignment between synthetic image-text pairs in VLM's feature space, we propose a contrastive distillation loss along with an image-text alignment constraint. To further combat in-distribution overfitting and enhance distillation performance with limited amount of generated data, we incorporate adaptive weight consolidation, utilizing Fisher information from these synthetic image-text pairs and achieving a better stability-plasticity balance. Extensive experiments demonstrate that our method consistently outperforms previous state-of-the-art approaches across various settings.
Authors: Hui Zhang, Zhiyang Wu, Qianqian Shangguan, Kang An
Abstract: Monocular visual localization plays a pivotal role in advanced driver assistance systems and autonomous driving by estimating a vehicle's ego-motion from a single pinhole camera. Nevertheless, conventional monocular visual odometry encoun-ters challenges in scale estimation due to the absence of depth information during projection. Previous methodologies, whether rooted in physical constraints or deep learning paradigms, con-tend with issues related to computational complexity and the management of dynamic objects. This study extends our prior research, presenting innovative strategies for ego-motion estima-tion and the selection of ground points. Striving for a nuanced equilibrium between computational efficiency and precision, we propose a hybrid method that leverages the SegNeXt model for real-time applications, encompassing both ego-motion estimation and ground point selection. Our methodology incorporates dy-namic object masks to eliminate unstable features and employs ground plane masks for meticulous triangulation. Furthermore, we exploit Geometry-constraint to delineate road regions for scale recovery. The integration of this approach with the mo-nocular version of ORB-SLAM3 culminates in the accurate esti-mation of a road model, a pivotal component in our scale recov-ery process. Rigorous experiments, conducted on the KITTI da-taset, systematically compare our method with existing monocu-lar visual odometry algorithms and contemporary scale recovery methodologies. The results undeniably confirm the superior ef-fectiveness of our approach, surpassing state-of-the-art visual odometry algorithms. Our source code is available at https://git hub.com/bFr0zNq/MVOSegScale.
URLs: https://git
Authors: Yifei Huang, Jilan Xu, Baoqi Pei, Yuping He, Guo Chen, Mingfang Zhang, Lijin Yang, Zheng Nie, Jinyao Liu, Guoshun Fan, Dechen Lin, Fang Fang, Kunpeng Li, Chang Yuan, Xinyuan Chen, Yaohui Wang, Yali Wang, Yu Qiao, Limin Wang
Abstract: We present Vinci, a vision-language system designed to provide real-time, comprehensive AI assistance on portable devices. At its core, Vinci leverages EgoVideo-VL, a novel model that integrates an egocentric vision foundation model with a large language model (LLM), enabling advanced functionalities such as scene understanding, temporal grounding, video summarization, and future planning. To enhance its utility, Vinci incorporates a memory module for processing long video streams in real time while retaining contextual history, a generation module for producing visual action demonstrations, and a retrieval module that bridges egocentric and third-person perspectives to provide relevant how-to videos for skill acquisition. Unlike existing systems that often depend on specialized hardware, Vinci is hardware-agnostic, supporting deployment across a wide range of devices, including smartphones and wearable cameras. In our experiments, we first demonstrate the superior performance of EgoVideo-VL on multiple public benchmarks, showcasing its vision-language reasoning and contextual understanding capabilities. We then conduct a series of user studies to evaluate the real-world effectiveness of Vinci, highlighting its adaptability and usability in diverse scenarios. We hope Vinci can establish a new framework for portable, real-time egocentric AI systems, empowering users with contextual and actionable insights. Including the frontend, backend, and models, all codes of Vinci are available at https://github.com/OpenGVLab/vinci.
Authors: Wonkwang Lee, Jongwon Jeong, Taehong Moon, Hyeon-Jong Kim, Jaehyeon Kim, Gunhee Kim, Byeong-Uk Lee
Abstract: Motion synthesis for diverse object categories holds great potential for 3D content creation but remains underexplored due to two key challenges: (1) the lack of comprehensive motion datasets that include a wide range of high-quality motions and annotations, and (2) the absence of methods capable of handling heterogeneous skeletal templates from diverse objects. To address these challenges, we contribute the following: First, we augment the Truebones Zoo dataset, a high-quality animal motion dataset covering over 70 species, by annotating it with detailed text descriptions, making it suitable for text-based motion synthesis. Second, we introduce rig augmentation techniques that generate diverse motion data while preserving consistent dynamics, enabling models to adapt to various skeletal configurations. Finally, we redesign existing motion diffusion models to dynamically adapt to arbitrary skeletal templates, enabling motion synthesis for a diverse range of objects with varying structures. Experiments show that our method learns to generate high-fidelity motions from textual descriptions for diverse and even unseen objects, setting a strong foundation for motion synthesis across diverse object categories and skeletal templates. Qualitative results are available on this link: t2m4lvo.github.io
Authors: Boseong Jeon
Abstract: In this report, I present an inpainting framework named \textit{ControlFill}, which involves training two distinct prompts: one for generating plausible objects within a designated mask (\textit{creation}) and another for filling the region by extending the background (\textit{removal}). During the inference stage, these learned embeddings guide a diffusion network that operates without requiring heavy text encoders. By adjusting the relative significance of the two prompts and employing classifier-free guidance, users can control the intensity of removal or creation. Furthermore, I introduce a method to spatially vary the intensity of guidance by assigning different scales to individual pixels.
Authors: Yecong Wan, Mingwen Shao, Yuanshuo Cheng, Wangmeng Zuo
Abstract: In this paper, we aim ambitiously for a realistic yet challenging problem, namely, how to reconstruct high-quality 3D scenes from sparse low-resolution views that simultaneously suffer from deficient perspectives and clarity. Whereas existing methods only deal with either sparse views or low-resolution observations, they fail to handle such hybrid and complicated scenarios. To this end, we propose a novel Sparse-view Super-resolution 3D Gaussian Splatting framework, dubbed S2Gaussian, that can reconstruct structure-accurate and detail-faithful 3D scenes with only sparse and low-resolution views. The S2Gaussian operates in a two-stage fashion. In the first stage, we initially optimize a low-resolution Gaussian representation with depth regularization and densify it to initialize the high-resolution Gaussians through a tailored Gaussian Shuffle Split operation. In the second stage, we refine the high-resolution Gaussians with the super-resolved images generated from both original sparse views and pseudo-views rendered by the low-resolution Gaussians. In which a customized blur-free inconsistency modeling scheme and a 3D robust optimization strategy are elaborately designed to mitigate multi-view inconsistency and eliminate erroneous updates caused by imperfect supervision. Extensive experiments demonstrate superior results and in particular establishing new state-of-the-art performances with more consistent geometry and finer details.
Authors: Lars Bredereke, Yale Hartmann, Tanja Schultz
Abstract: Object tracking is a key challenge of computer vision with various applications that all require different architectures. Most tracking systems have limitations such as constraining all movement to a 2D plane and they often track only one object. In this paper, we present a new modular pipeline that calculates 3D trajectories of multiple objects. It is adaptable to various settings where multiple time-synced and stationary cameras record moving objects, using off the shelf webcams. Our pipeline was tested on the Table Setting Dataset, where participants are recorded with various sensors as they set a table with tableware objects. We need to track these manipulated objects, using 6 rgb webcams. Challenges include: Detecting small objects in 9.874.699 camera frames, determining camera poses, discriminating between nearby and overlapping objects, temporary occlusions, and finally calculating a 3D trajectory using the right subset of an average of 11.12.456 pixel coordinates per 3-minute trial. We implement a robust pipeline that results in accurate trajectories with covariance of x,y,z-position as a confidence metric. It deals dynamically with appearing and disappearing objects, instantiating new Extended Kalman Filters. It scales to hundreds of table-setting trials with very little human annotation input, even with the camera poses of each trial unknown. The code is available at https://github.com/LarsBredereke/object_tracking
Authors: Inseo Lee, Youngyoon Choi, Joonseok Lee
Abstract: Implicit Neural Representation for Videos (NeRV) has introduced a novel paradigm for video representation and compression, outperforming traditional codecs. As model size grows, however, slow encoding and decoding speed and high memory consumption hinder its application in practice. To address these limitations, we propose a new video representation and compression method based on 2D Gaussian Splatting to efficiently handle video data. Our proposed deformable 2D Gaussian Splatting dynamically adapts the transformation of 2D Gaussians at each frame, significantly reducing memory cost. Equipped with a multi-plane-based spatiotemporal encoder and a lightweight decoder, it predicts changes in color, coordinates, and shape of initialized Gaussians, given the time step. By leveraging temporal gradients, our model effectively captures temporal redundancy at negligible cost, significantly enhancing video representation efficiency. Our method reduces GPU memory usage by up to 78.4%, and significantly expedites video processing, achieving 5.5x faster training and 12.5x faster decoding compared to the state-of-the-art NeRV methods.
Authors: Shen Zhang, Yaning Tan, Siyuan Liang, Linze Li, Ge Wu, Yuhao Chen, Shuheng Li, Zhenyu Zhao, Caihua Chen, Jiajun Liang, Yao Tang
Abstract: Diffusion transformers(DiTs) struggle to generate images at resolutions higher than their training resolutions. The primary obstacle is that the explicit positional encodings(PE), such as RoPE, need extrapolation which degrades performance when the inference resolution differs from training. In this paper, we propose a Length-Extrapolatable Diffusion Transformer(LEDiT), a simple yet powerful architecture to overcome this limitation. LEDiT needs no explicit PEs, thereby avoiding extrapolation. The key innovations of LEDiT are introducing causal attention to implicitly impart global positional information to tokens, while enhancing locality to precisely distinguish adjacent tokens. Experiments on 256x256 and 512x512 ImageNet show that LEDiT can scale the inference resolution to 512x512 and 1024x1024, respectively, while achieving better image quality compared to current state-of-the-art length extrapolation methods(NTK-aware, YaRN). Moreover, LEDiT achieves strong extrapolation performance with just 100K steps of fine-tuning on a pretrained DiT, demonstrating its potential for integration into existing text-to-image DiTs.
Authors: Kim Kiehn, Albin Ahlb\"ack, Kathl\'en Kohn
Abstract: We completely classify all minimal problems for Structure-from-Motion (SfM) where arrangements of points and lines are fully observed by multiple uncalibrated pinhole cameras. We find 291 minimal problems, 73 of which have unique solutions and can thus be solved linearly. Two of the linear problems allow an arbitrary number of views, while all other minimal problems have at most 9 cameras. All minimal problems have at most 7 points and at most 12 lines. We compute the number of solutions of each minimal problem, as this gives a measurement of the problem's intrinsic difficulty, and find that these number are relatively low (e.g., when comparing with minimal problems for calibrated cameras). Finally, by exploring stabilizer subgroups of subarrangements, we develop a geometric and systematic way to 1) factorize minimal problems into smaller problems, 2) identify minimal problems in underconstrained problems, and 3) formally prove non-minimality.
Authors: Chanda Grover Kamra, Indra Deep Mastan, Debayan Gupta
Abstract: We propose ObjMST, an object-focused multimodal style transfer framework that provides separate style supervision for salient objects and surrounding elements while addressing alignment issues in multimodal representation learning. Existing image-text multimodal style transfer methods face the following challenges: (1) generating non-aligned and inconsistent multimodal style representations; and (2) content mismatch, where identical style patterns are applied to both salient objects and their surrounding elements. Our approach mitigates these issues by: (1) introducing a Style-Specific Masked Directional CLIP Loss, which ensures consistent and aligned style representations for both salient objects and their surroundings; and (2) incorporating a salient-to-key mapping mechanism for stylizing salient objects, followed by image harmonization to seamlessly blend the stylized objects with their environment. We validate the effectiveness of ObjMST through experiments, using both quantitative metrics and qualitative visual evaluations of the stylized outputs. Our code is available at: https://github.com/chandagrover/ObjMST.
Authors: Peng Xu, Zhiyu Xiang, Jingyun Fu, Tianyu Pu, Hanzhi Zhong, Eryun Liu
Abstract: Despite the significant advances in domain generalized stereo matching, existing methods still exhibit domain-specific preferences when transferring from synthetic to real domains, hindering their practical applications in complex and diverse scenarios. The probability distributions predicted by the stereo network naturally encode rich similarity and uncertainty information. Inspired by this observation, we propose to extract these two types of dark knowledge from the pre-trained network to model intuitive multi-modal ground-truth distributions for both edge and non-edge regions. To mitigate the inherent domain preferences of a single network, we adopt network ensemble and further distinguish between objective and biased knowledge in the Laplace parameter space. Finally, the objective knowledge and the original disparity labels are jointly modeled as a mixture of Laplacians to provide fine-grained supervision for the stereo network training. Extensive experiments demonstrate that: 1) Our method is generic and effectively improves the generalization of existing networks. 2) PCWNet with our method achieves the state-of-the-art generalization performance on both KITTI 2015 and 2012 datasets. 3) Our method outperforms existing methods in comprehensive ranking across four popular real-world datasets.
Authors: Yihao Huang, Xin Luo, Qing Guo, Felix Juefei-Xu, Xiaojun Jia, Weikai Miao, Geguang Pu, Yang Liu
Abstract: The advent of local continuous image function (LIIF) has garnered significant attention for arbitrary-scale super-resolution (SR) techniques. However, while the vulnerabilities of fixed-scale SR have been assessed, the robustness of continuous representation-based arbitrary-scale SR against adversarial attacks remains an area warranting further exploration. The elaborately designed adversarial attacks for fixed-scale SR are scale-dependent, which will cause time-consuming and memory-consuming problems when applied to arbitrary-scale SR. To address this concern, we propose a simple yet effective ``scale-invariant'' SR adversarial attack method with good transferability, termed SIAGT. Specifically, we propose to construct resource-saving attacks by exploiting finite discrete points of continuous representation. In addition, we formulate a coordinate-dependent loss to enhance the cross-model transferability of the attack. The attack can significantly deteriorate the SR images while introducing imperceptible distortion to the targeted low-resolution (LR) images. Experiments carried out on three popular LIIF-based SR approaches and four classical SR datasets show remarkable attack performance and transferability of SIAGT.
Authors: Harry J. F. Owen, Matthew J. A. Allen, Stuart W. D. Grieve, Phill Wilkes, Emily R. Lines
Abstract: Point clouds from Terrestrial Laser Scanning (TLS) are an increasingly popular source of data for studying plant structure and function but typically require extensive manual processing to extract ecologically important information. One key task is the accurate semantic segmentation of different plant material within point clouds, particularly wood and leaves, which is required to understand plant productivity, architecture and physiology. Existing automated semantic segmentation methods are primarily developed for single ecosystem types, and whilst they show good accuracy for biomass assessment from the trunk and large branches, often perform less well within the crown. In this study, we demonstrate a new framework that uses a deep learning architecture newly developed from PointNet and pointNEXT for processing 3D point clouds to provide a reliable semantic segmentation of wood and leaf in TLS point clouds from the tree base to branch tips, trained on data from diverse mature European forests. Our model uses meticulously labelled data combined with voxel-based sampling, neighbourhood rescaling, and a novel gated reflectance integration module embedded throughout the feature extraction layers. We evaluate its performance across open datasets from boreal, temperate, Mediterranean and tropical regions, encompassing diverse ecosystem types and sensor characteristics. Our results show consistent outperformance against the most widely used PointNet based approach for leaf/wood segmentation on our high-density TLS dataset collected across diverse mixed forest plots across all major biomes in Europe. We also find consistently strong performance tested on others open data from China, Eastern Cameroon, Germany and Finland, collected using both time-of-flight and phase-shift sensors, showcasing the transferability of our model to a wide range of ecosystems and sensors.
Authors: Vittorio Pippi, Matthieu Guillaumin, Silvia Cascianelli, Rita Cucchiara, Maximilian Jaritz, Loris Bazzani
Abstract: Large Multimodal Models (LMMs) are powerful tools that are capable of reasoning and understanding multimodal information beyond text and language. Despite their entrenched impact, the development of LMMs is hindered by the higher computational requirements compared to their unimodal counterparts. One of the main causes of this is the large amount of tokens needed to encode the visual input, which is especially evident for multi-image multimodal tasks. Recent approaches to reduce visual tokens depend on the visual encoder architecture, require fine-tuning the LLM to maintain the performance, and only consider single-image scenarios. To address these limitations, we propose ToFu, a visual encoder-agnostic, training-free Token Fusion strategy that combines redundant visual tokens of LMMs for high-resolution, multi-image, tasks. The core intuition behind our method is straightforward yet effective: preserve distinctive tokens while combining similar ones. We achieve this by sequentially examining visual tokens and deciding whether to merge them with others or keep them as separate entities. We validate our approach on the well-established LLaVA-Interleave Bench, which covers challenging multi-image tasks. In addition, we push to the extreme our method by testing it on a newly-created benchmark, ComPairs, focused on multi-image comparisons where a larger amount of images and visual tokens are inputted to the LMMs. Our extensive analysis, considering several LMM architectures, demonstrates the benefits of our approach both in terms of efficiency and performance gain.
Authors: Xuerui Zhang
Abstract: Small targets are particularly difficult to detect due to their low pixel count, complex backgrounds, and varying shooting angles, which make it hard for models to extract effective features. While some large-scale models offer high accuracy, their long inference times make them unsuitable for real-time deployment on edge devices. On the other hand, models designed for low computational power often suffer from poor detection accuracy. This paper focuses on small target detection and explores methods for object detection under low computational constraints. Building on the YOLOv8 model, we propose a new network architecture called FDM-YOLO. Our research includes the following key contributions: We introduce FDM-YOLO by analyzing the output of the YOLOv8 detection head. We add a highresolution layer and remove the large target detection layer to better handle small targets. Based on PConv, we propose a lightweight network structure called Fast-C2f, which is integrated into the PAN module of the model. To mitigate the accuracy loss caused by model lightweighting, we employ dynamic upsampling (Dysample) and a lightweight EMA attention mechanism.The FDM-YOLO model was validated on the Visdrone dataset, achieving a 38% reduction in parameter count and improving the Map0.5 score from 38.4% to 42.5%, all while maintaining nearly the same inference speed. This demonstrates the effectiveness of our approach in balancing accuracy and efficiency for edge device deployment.
Authors: Chao Wang, Weiwei Fu, Yang Zhou
Abstract: Vision-language models (VLMs) have achieved remarkable advancements, capitalizing on the impressive capabilities of large language models (LLMs) across diverse tasks. Despite this, a critical challenge known as hallucination occurs when models overconfidently describe objects or attributes absent from the image, a problem exacerbated by the tendency of VLMs to rely on linguistic priors. This limitation reduces model reliability in high-stakes applications. In this work, we have observed the characteristic of logits' continuity consistency enhancement and introduced a straightforward and efficient method, Cross-Temporal Prediction Connection (TPC), designed to enhance the semantic consistency of logits by connecting them temporally across timesteps. TPC amplifies information flow and improves coherence, effectively reducing hallucination. Extensive experiments show that TPC surpasses existing representatives, delivering superior performance in both accuracy and efficiency while maintaining robustness in open-ended text generation tasks.
Authors: Hongyeob Kim, Inyoung Jung, Dayoon Suh, Youjia Zhang, Sangmin Lee, Sungeun Hong
Abstract: Audio-Visual Question Answering (AVQA) requires not only question-based multimodal reasoning but also precise temporal grounding to capture subtle dynamics for accurate prediction. However, existing methods mainly use question information implicitly, limiting focus on question-specific details. Furthermore, most studies rely on uniform frame sampling, which can miss key question-relevant frames. Although recent Top-K frame selection methods aim to address this, their discrete nature still overlooks fine-grained temporal details. This paper proposes \textbf{QA-TIGER}, a novel framework that explicitly incorporates question information and models continuous temporal dynamics. Our key idea is to use Gaussian-based modeling to adaptively focus on both consecutive and non-consecutive frames based on the question, while explicitly injecting question information and applying progressive refinement. We leverage a Mixture of Experts (MoE) to flexibly implement multiple Gaussian models, activating temporal experts specifically tailored to the question. Extensive experiments on multiple AVQA benchmarks show that QA-TIGER consistently achieves state-of-the-art performance. Code is available at https://github.com/AIM-SKKU/QA-TIGER
Authors: Edoardo Bianchi, Oswald Lanz
Abstract: This paper introduces Gate-Shift-Pose, an enhanced version of Gate-Shift-Fuse networks, designed for athlete fall classification in figure skating by integrating skeleton pose data alongside RGB frames. We evaluate two fusion strategies: early-fusion, which combines RGB frames with Gaussian heatmaps of pose keypoints at the input stage, and late-fusion, which employs a multi-stream architecture with attention mechanisms to combine RGB and pose features. Experiments on the FR-FS dataset demonstrate that Gate-Shift-Pose significantly outperforms the RGB-only baseline, improving accuracy by up to 40% with ResNet18 and 20% with ResNet50. Early-fusion achieves the highest accuracy (98.08%) with ResNet50, leveraging the model's capacity for effective multimodal integration, while late-fusion is better suited for lighter backbones like ResNet18. These results highlight the potential of multimodal architectures for sports action recognition and the critical role of skeleton pose information in capturing complex motion patterns.
Authors: Yanqing Shen, Turcan Tuna, Marco Hutter, Cesar Cadena, Nanning Zheng
Abstract: Place recognition is essential to maintain global consistency in large-scale localization systems. While research in urban environments has progressed significantly using LiDARs or cameras, applications in natural forest-like environments remain largely under-explored. Furthermore, forests present particular challenges due to high self-similarity and substantial variations in vegetation growth over time. In this work, we propose a robust LiDAR-based place recognition method for natural forests, ForestLPR. We hypothesize that a set of cross-sectional images of the forest's geometry at different heights contains the information needed to recognize revisiting a place. The cross-sectional images are represented by \ac{bev} density images of horizontal slices of the point cloud at different heights. Our approach utilizes a visual transformer as the shared backbone to produce sets of local descriptors and introduces a multi-BEV interaction module to attend to information at different heights adaptively. It is followed by an aggregation layer that produces a rotation-invariant place descriptor. We evaluated the efficacy of our method extensively on real-world data from public benchmarks as well as robotic datasets and compared it against the state-of-the-art (SOTA) methods. The results indicate that ForestLPR has consistently good performance on all evaluations and achieves an average increase of 7.38\% and 9.11\% on Recall@1 over the closest competitor on intra-sequence loop closure detection and inter-sequence re-localization, respectively, validating our hypothesis
Authors: Maxime Di Folco, Emily Chan, Marta Hasny, Cosmin I. Bercea, Julia A. Schnabel
Abstract: General-purpose AI models, particularly those designed for text and vision, demonstrate impressive versatility across a wide range of deep-learning tasks. However, they often underperform in specialised domains like medical imaging, where domain-specific solutions or alternative knowledge transfer approaches are typically required. Recent studies have noted that general-purpose models can exhibit similar latent spaces when processing semantically related data, although this alignment does not occur naturally. Building on this insight, it has been shown that applying a simple transformation - at most affine - estimated from a subset of semantically corresponding samples, known as anchors, enables model stitching across diverse training paradigms, architectures, and modalities. In this paper, we explore how semantic alignment - estimating transformations between anchors - can bridge general-purpose AI with specialised medical knowledge. Using multiple public chest X-ray datasets, we demonstrate that model stitching across model architectures allows general models to integrate domain-specific knowledge without additional training, leading to improved performance on medical tasks. Furthermore, we introduce a novel zero-shot classification approach for unimodal vision encoders that leverages semantic alignment across modalities. Our results show that our method not only outperforms general multimodal models but also approaches the performance levels of fully trained, medical-specific multimodal solutions
Authors: Benjamin Billot, Ramya Muthukrishnan, Esra Abaci-Turk, Ellen P. Grant, Nicholas Ayache, Herv\'e Delingette, Polina Golland
Abstract: Unsupervised registration strategies bypass requirements in ground truth transforms or segmentations by optimising similarity metrics between fixed and moved volumes. Among these methods, a recent subclass of approaches based on unsupervised keypoint detection stand out as very promising for interpretability. Specifically, these methods train a network to predict feature maps for fixed and moving images, from which explainable centres of mass are computed to obtain point clouds, that are then aligned in closed-form. However, the features returned by the network often yield spatially diffuse patterns that are hard to interpret, thus undermining the purpose of keypoint-based registration. Here, we propose a three-fold loss to regularise the spatial distribution of the features. First, we use the KL divergence to model features as point spread functions that we interpret as probabilistic keypoints. Then, we sharpen the spatial distributions of these features to increase the precision of the detected landmarks. Finally, we introduce a new repulsive loss across keypoints to encourage spatial diversity. Overall, our loss considerably improves the interpretability of the features, which now correspond to precise and anatomically meaningful landmarks. We demonstrate our three-fold loss in foetal rigid motion tracking and brain MRI affine registration tasks, where it not only outperforms state-of-the-art unsupervised strategies, but also bridges the gap with state-of-the-art supervised methods. Our code is available at https://github.com/BenBillot/spatial_regularisation.
Authors: Yu-Hsi Chen, Chin-Tien Wu
Abstract: Optical flow is a fundamental technique for motion estimation, widely applied in video stabilization, interpolation, and object tracking. Recent advancements in artificial intelligence (AI) have enabled deep learning models to leverage optical flow as an important feature for motion analysis. However, traditional optical flow methods rely on restrictive assumptions, such as brightness constancy and slow motion constraints, limiting their effectiveness in complex scenes. Deep learning-based approaches require extensive training on large domain-specific datasets, making them computationally demanding. Furthermore, optical flow is typically visualized in the HSV color space, which introduces nonlinear distortions when converted to RGB and is highly sensitive to noise, degrading motion representation accuracy. These limitations inherently constrain the performance of downstream models, potentially hindering object tracking and motion analysis tasks. To address these challenges, we propose Reynolds flow, a novel training-free flow estimation inspired by the Reynolds transport theorem, offering a principled approach to modeling complex motion dynamics. Beyond the conventional HSV-based visualization, denoted ReynoldsFlow, we introduce an alternative representation, ReynoldsFlow+, designed to improve flow visualization. We evaluate ReynoldsFlow and ReynoldsFlow+ across three video-based benchmarks: tiny object detection on UAVDB, infrared object detection on Anti-UAV, and pose estimation on GolfDB. Experimental results demonstrate that networks trained with ReynoldsFlow+ achieve state-of-the-art (SOTA) performance, exhibiting improved robustness and efficiency across all tasks.
Authors: Zhihao Shi, Dong Huo, Yuhongze Zhou, Kejia Yin, Yan Min, Juwei Lu, Xinxin Zuo
Abstract: Current 3D inpainting and object removal methods are largely limited to front-facing scenes, facing substantial challenges when applied to diverse, "unconstrained" scenes where the camera orientation and trajectory are unrestricted. To bridge this gap, we introduce a novel approach that produces inpainted 3D scenes with consistent visual quality and coherent underlying geometry across both front-facing and unconstrained scenes. Specifically, we propose a robust 3D inpainting pipeline that incorporates geometric priors and a multi-view refinement network trained via test-time adaptation, building on a pre-trained image inpainting model. Additionally, we develop a novel inpainting mask detection technique to derive targeted inpainting masks from object masks, boosting the performance in handling unconstrained scenes. To validate the efficacy of our approach, we create a challenging and diverse benchmark that spans a wide range of scenes. Comprehensive experiments demonstrate that our proposed method substantially outperforms existing state-of-the-art approaches.
Authors: Sunghyun Ahn, Youngwan Jo, Kijung Lee, Sein Kwon, Inpyo Hong, Sanghyun Park
Abstract: Video anomaly detection (VAD) is crucial for video analysis and surveillance in computer vision. However, existing VAD models rely on learned normal patterns, which makes them difficult to apply to diverse environments. Consequently, users should retrain models or develop separate AI models for new environments, which requires expertise in machine learning, high-performance hardware, and extensive data collection, limiting the practical usability of VAD. To address these challenges, this study proposes customizable video anomaly detection (C-VAD) technique and the AnyAnomaly model. C-VAD considers user-defined text as an abnormal event and detects frames containing a specified event in a video. We effectively implemented AnyAnomaly using a context-aware visual question answering without fine-tuning the large vision language model. To validate the effectiveness of the proposed model, we constructed C-VAD datasets and demonstrated the superiority of AnyAnomaly. Furthermore, our approach showed competitive performance on VAD benchmark datasets, achieving state-of-the-art results on the UBnormal dataset and outperforming other methods in generalization across all datasets. Our code is available online at github.com/SkiddieAhn/Paper-AnyAnomaly.
Authors: Jiageng Zhong, Qi Zhou, Ming Li, Armin Gruen, Xuan Liao
Abstract: Low-overlap aerial imagery poses significant challenges to traditional photogrammetric methods, which rely heavily on high image overlap to produce accurate and complete mapping products. In this study, we propose a novel workflow based on monocular depth estimation to address the limitations of conventional techniques. Our method leverages tie points obtained from aerial triangulation to establish a relationship between monocular depth and metric depth, thus transforming the original depth map into a metric depth map, enabling the generation of dense depth information and the comprehensive reconstruction of the scene. For the experiments, a high-overlap drone dataset containing 296 images is processed using Metashape to generate depth maps and DSMs as ground truth. Subsequently, we create a low-overlap dataset by selecting 20 images for experimental evaluation. Results demonstrate that while the recovered depth maps and resulting DSMs achieve meter-level accuracy, they provide significantly better completeness compared to traditional methods, particularly in regions covered by single images. This study showcases the potential of monocular depth estimation in low-overlap aerial photogrammetry.
Authors: Matias Cosarinsky, Ramiro Billot, Lucas Mansilla, Gabriel Gimenez, Nicolas Gaggi\'on, Guanghui Fu, Enzo Ferrante
Abstract: Assessing the quality of automatic image segmentation is crucial in clinical practice, but often very challenging due to the limited availability of ground truth annotations. In this paper, we introduce In-Context Reverse Classification Accuracy (In-Context RCA), a novel framework for automatically estimating segmentation quality in the absence of ground-truth annotations. By leveraging recent in-context learning segmentation models and incorporating retrieval-augmentation techniques to select the most relevant reference images, our approach enables efficient quality estimation with minimal reference data. Validated across diverse medical imaging modalities, our method demonstrates robust performance and computational efficiency, offering a promising solution for automated quality control in clinical workflows, where fast and reliable segmentation assessment is essential. The code is available at https://github.com/mcosarinsky/In-Context-RCA.
Authors: Kai Luo, Hao Shi, Sheng Wu, Fei Teng, Mengfei Duan, Chang Huang, Yuhang Wang, Kaiwei Wang, Kailun Yang
Abstract: Panoramic imagery, with its 360{\deg} field of view, offers comprehensive information to support Multi-Object Tracking (MOT) in capturing spatial and temporal relationships of surrounding objects. However, most MOT algorithms are tailored for pinhole images with limited views, impairing their effectiveness in panoramic settings. Additionally, panoramic image distortions, such as resolution loss, geometric deformation, and uneven lighting, hinder direct adaptation of existing MOT methods, leading to significant performance degradation. To address these challenges, we propose OmniTrack, an omnidirectional MOT framework that incorporates Tracklet Management to introduce temporal cues, FlexiTrack Instances for object localization and association, and the CircularStatE Module to alleviate image and geometric distortions. This integration enables tracking in large field-of-view scenarios, even under rapid sensor motion. To mitigate the lack of panoramic MOT datasets, we introduce the QuadTrack dataset--a comprehensive panoramic dataset collected by a quadruped robot, featuring diverse challenges such as wide fields of view, intense motion, and complex environments. Extensive experiments on the public JRDB dataset and the newly introduced QuadTrack benchmark demonstrate the state-of-the-art performance of the proposed framework. OmniTrack achieves a HOTA score of 26.92% on JRDB, representing an improvement of 3.43%, and further achieves 23.45% on QuadTrack, surpassing the baseline by 6.81%. The dataset and code will be made publicly available at https://github.com/xifen523/OmniTrack.
Authors: Qing Zhou, Tao Yang, Junyu Gao, Weiping Ni, Junzheng Wu, Qi Wang
Abstract: Remote Sensing Image Captioning (RSIC) is a cross-modal field bridging vision and language, aimed at automatically generating natural language descriptions of features and scenes in remote sensing imagery. Despite significant advances in developing sophisticated methods and large-scale datasets for training vision-language models (VLMs), two critical challenges persist: the scarcity of non-English descriptive datasets and the lack of multilingual capability evaluation for models. These limitations fundamentally impede the progress and practical deployment of RSIC, particularly in the era of large VLMs. To address these challenges, this paper presents several significant contributions to the field. First, we introduce and analyze BRSIC (Bilingual Remote Sensing Image Captioning), a comprehensive bilingual dataset that enriches three established English RSIC datasets with Chinese descriptions, encompassing 13,634 images paired with 68,170 bilingual captions. Building upon this foundation, we develop a systematic evaluation framework that addresses the prevalent inconsistency in evaluation protocols, enabling rigorous assessment of model performance through standardized retraining procedures on BRSIC. Furthermore, we present an extensive empirical study of eight state-of-the-art large vision-language models (LVLMs), examining their capabilities across multiple paradigms including zero-shot inference, supervised fine-tuning, and multi-lingual training. This comprehensive evaluation provides crucial insights into the strengths and limitations of current LVLMs in handling multilingual remote sensing tasks. Additionally, our cross-dataset transfer experiments reveal interesting findings. The code and data will be available at https://github.com/mrazhou/BRSIC.
Authors: Aoxiong Yin, Kai Shen, Yichong Leng, Xu Tan, Xinyu Zhou, Juncheng Li, Siliang Tang
Abstract: Recent advancements in text-to-video (T2V) generation have been driven by two competing paradigms: autoregressive language models and diffusion models. However, each paradigm has intrinsic limitations: language models struggle with visual quality and error accumulation, while diffusion models lack semantic understanding and causal modeling. In this work, we propose LanDiff, a hybrid framework that synergizes the strengths of both paradigms through coarse-to-fine generation. Our architecture introduces three key innovations: (1) a semantic tokenizer that compresses 3D visual features into compact 1D discrete representations through efficient semantic compression, achieving a $\sim$14,000$\times$ compression ratio; (2) a language model that generates semantic tokens with high-level semantic relationships; (3) a streaming diffusion model that refines coarse semantics into high-fidelity videos. Experiments show that LanDiff, a 5B model, achieves a score of 85.43 on the VBench T2V benchmark, surpassing the state-of-the-art open-source models Hunyuan Video (13B) and other commercial models such as Sora, Keling, and Hailuo. Furthermore, our model also achieves state-of-the-art performance in long video generation, surpassing other open-source models in this field. Our demo can be viewed at https://landiff.github.io/.
Authors: Hong Liu, Haosen Yang, Evi M. C. Huijben, Mark Schuiveling, Ruisheng Su, Josien P. W. Pluim, Mitko Veta
Abstract: Tumor segmentation plays a critical role in histopathology, but it requires costly, fine-grained image-mask pairs annotated by pathologists. Thus, synthesizing histopathology data to expand the dataset is highly desirable. Previous works suffer from inaccuracies and limited diversity in image-mask pairs, both of which affect training segmentation, particularly in small-scale datasets and the inherently complex nature of histopathology images. To address this challenge, we propose PathoPainter, which reformulates image-mask pair generation as a tumor inpainting task. Specifically, our approach preserves the background while inpainting the tumor region, ensuring precise alignment between the generated image and its corresponding mask. To enhance dataset diversity while maintaining biological plausibility, we incorporate a sampling mechanism that conditions tumor inpainting on regional embeddings from a different image. Additionally, we introduce a filtering strategy to exclude uncertain synthetic regions, further improving the quality of the generated data. Our comprehensive evaluation spans multiple datasets featuring diverse tumor types and various training data scales. As a result, segmentation improved significantly with our synthetic data, surpassing existing segmentation data synthesis approaches, e.g., 75.69% -> 77.69% on CAMELYON16. The code is available at https://github.com/HongLiuuuuu/PathoPainter.
Authors: Aishik Konwer, Zhijian Yang, Erhan Bas, Cao Xiao, Prateek Prasanna, Parminder Bhatia, Taha Kass-Hout
Abstract: Foundational models such as the Segment Anything Model (SAM) are gaining traction in medical imaging segmentation, supporting multiple downstream tasks. However, such models are supervised in nature, still relying on large annotated datasets or prompts supplied by experts. Conventional techniques such as active learning to alleviate such limitations are limited in scope and still necessitate continuous human involvement and complex domain knowledge for label refinement or establishing reward ground truth. To address these challenges, we propose an enhanced Segment Anything Model (SAM) framework that utilizes annotation-efficient prompts generated in a fully unsupervised fashion, while still capturing essential semantic, location, and shape information through contrastive language-image pretraining and visual question answering. We adopt the direct preference optimization technique to design an optimal policy that enables the model to generate high-fidelity segmentations with simple ratings or rankings provided by a virtual annotator simulating the human annotation process. State-of-the-art performance of our framework in tasks such as lung segmentation, breast tumor segmentation, and organ segmentation across various modalities, including X-ray, ultrasound, and abdominal CT, justifies its effectiveness in low-annotation data scenarios.
Authors: Yuqi Hu, Longguang Wang, Xian Liu, Ling-Hao Chen, Yuwei Guo, Yukai Shi, Ce Liu, Anyi Rao, Zeyu Wang, Hui Xiong
Abstract: Understanding and replicating the real world is a critical challenge in Artificial General Intelligence (AGI) research. To achieve this, many existing approaches, such as world models, aim to capture the fundamental principles governing the physical world, enabling more accurate simulations and meaningful interactions. However, current methods often treat different modalities, including 2D (images), videos, 3D, and 4D representations, as independent domains, overlooking their interdependencies. Additionally, these methods typically focus on isolated dimensions of reality without systematically integrating their connections. In this survey, we present a unified survey for multimodal generative models that investigate the progression of data dimensionality in real-world simulation. Specifically, this survey starts from 2D generation (appearance), then moves to video (appearance+dynamics) and 3D generation (appearance+geometry), and finally culminates in 4D generation that integrate all dimensions. To the best of our knowledge, this is the first attempt to systematically unify the study of 2D, video, 3D and 4D generation within a single framework. To guide future research, we provide a comprehensive review of datasets, evaluation metrics and future directions, and fostering insights for newcomers. This survey serves as a bridge to advance the study of multimodal generative models and real-world simulation within a unified framework.
Authors: Tengfei Zhang, Ziheng Zhao, Chaoyi Wu, Xiao Zhou, Ya Zhang, Yangfeng Wang, Weidi Xie
Abstract: Developing advanced medical imaging retrieval systems is challenging due to the varying definitions of `similar images' across different medical contexts. This challenge is compounded by the lack of large-scale, high-quality medical imaging retrieval datasets and benchmarks. In this paper, we propose a novel methodology that leverages dense radiology reports to define image-wise similarity ordering at multiple granularities in a scalable and fully automatic manner. Using this approach, we construct two comprehensive medical imaging retrieval datasets: MIMIC-IR for Chest X-rays and CTRATE-IR for CT scans, providing detailed image-image ranking annotations conditioned on diverse anatomical structures. Furthermore, we develop two retrieval systems, RadIR-CXR and model-ChestCT, which demonstrate superior performance in traditional image-image and image-report retrieval tasks. These systems also enable flexible, effective image retrieval conditioned on specific anatomical structures described in text, achieving state-of-the-art results on 77 out of 78 metrics.
Authors: Mary Aiyetigbo, Wanqi Yuan, Feng Luo, Nianyi Li
Abstract: We present a novel approach for super-resolution that utilizes implicit neural representation (INR) to effectively reconstruct and enhance low-resolution videos and images. By leveraging the capacity of neural networks to implicitly encode spatial and temporal features, our method facilitates high-resolution reconstruction using only low-resolution inputs and a 3D high-resolution grid. This results in an efficient solution for both image and video super-resolution. Our proposed method, SR-INR, maintains consistent details across frames and images, achieving impressive temporal stability without relying on the computationally intensive optical flow or motion estimation typically used in other video super-resolution techniques. The simplicity of our approach contrasts with the complexity of many existing methods, making it both effective and efficient. Experimental evaluations show that SR-INR delivers results on par with or superior to state-of-the-art super-resolution methods, while maintaining a more straightforward structure and reduced computational demands. These findings highlight the potential of implicit neural representations as a powerful tool for reconstructing high-quality, temporally consistent video and image signals from low-resolution data.
Authors: Emanuele Bugliarello, Anurag Arnab, Roni Paiss, Pieter-Jan Kindermans, Cordelia Schmid
Abstract: High-quality benchmarks are crucial for driving progress in machine learning research. However, despite the growing interest in video generation, there is no comprehensive dataset to evaluate human generation. Humans can perform a wide variety of actions and interactions, but existing datasets, like TikTok and TED-Talks, lack the diversity and complexity to fully capture the capabilities of video generation models. We close this gap by introducing `What Are You Doing?' (WYD): a new benchmark for fine-grained evaluation of controllable image-to-video generation of humans. WYD consists of 1{,}544 captioned videos that have been meticulously collected and annotated with 56 fine-grained categories. These allow us to systematically measure performance across 9 aspects of human generation, including actions, interactions and motion. We also propose and validate automatic metrics that leverage our annotations and better capture human evaluations. Equipped with our dataset and metrics, we perform in-depth analyses of seven state-of-the-art models in controllable image-to-video generation, showing how WYD provides novel insights about the capabilities of these models. We release our data and code to drive forward progress in human video generation modeling at https://github.com/google-deepmind/wyd-benchmark.
Authors: Riccardo De Monte, Davide Dalle Pezze, Gian Antonio Susto
Abstract: Real-time object detectors like YOLO achieve exceptional performance when trained on large datasets for multiple epochs. However, in real-world scenarios where data arrives incrementally, neural networks suffer from catastrophic forgetting, leading to a loss of previously learned knowledge. To address this, prior research has explored strategies for Class Incremental Learning (CIL) in Continual Learning for Object Detection (CLOD), with most approaches focusing on two-stage object detectors. However, existing work suggests that Learning without Forgetting (LwF) may be ineffective for one-stage anchor-free detectors like YOLO due to noisy regression outputs, which risk transferring corrupted knowledge. In this work, we introduce YOLO LwF, a self-distillation approach tailored for YOLO-based continual object detection. We demonstrate that when coupled with a replay memory, YOLO LwF significantly mitigates forgetting. Compared to previous approaches, it achieves state-of-the-art performance, improving mAP by +2.1% and +2.9% on the VOC and COCO benchmarks, respectively.
Authors: Aditya Prashant Naidu, Hem Gosalia, Ishaan Gakhar, Shaurya Singh Rathore, Krish Didwania, Ujjwal Verma
Abstract: Although advances in deep learning and aerial surveillance technology are improving wildlife conservation efforts, complex and erratic environmental conditions still pose a problem, requiring innovative solutions for cost-effective small animal detection. This work introduces DEAL-YOLO, a novel approach that improves small object detection in Unmanned Aerial Vehicle (UAV) images by using multi-objective loss functions like Wise IoU (WIoU) and Normalized Wasserstein Distance (NWD), which prioritize pixels near the centre of the bounding box, ensuring smoother localization and reducing abrupt deviations. Additionally, the model is optimized through efficient feature extraction with Linear Deformable (LD) convolutions, enhancing accuracy while maintaining computational efficiency. The Scaled Sequence Feature Fusion (SSFF) module enhances object detection by effectively capturing inter-scale relationships, improving feature representation, and boosting metrics through optimized multiscale fusion. Comparison with baseline models reveals high efficacy with up to 69.5\% fewer parameters compared to vanilla Yolov8-N, highlighting the robustness of the proposed modifications. Through this approach, our paper aims to facilitate the detection of endangered species, animal population analysis, habitat monitoring, biodiversity research, and various other applications that enrich wildlife conservation efforts. DEAL-YOLO employs a two-stage inference paradigm for object detection, refining selected regions to improve localization and confidence. This approach enhances performance, especially for small instances with low objectness scores.
Authors: Mengdi Wang, Efe Bozkir, Enkelejda Kasneci
Abstract: Iris texture is widely regarded as a gold standard biometric modality for authentication and identification. The demand for robust iris recognition methods, coupled with growing security and privacy concerns regarding iris attacks, has escalated recently. Inspired by neural style transfer, an advanced technique that leverages neural networks to separate content and style features, we hypothesize that iris texture's style features provide a reliable foundation for recognition and are more resilient to variations like rotation and perspective shifts than traditional approaches. Our experimental results support this hypothesis, showing a significantly higher classification accuracy compared to conventional features. Further, we propose using neural style transfer to mask identifiable iris style features, ensuring the protection of sensitive biometric information while maintaining the utility of eye images for tasks like eye segmentation and gaze estimation. This work opens new avenues for iris-oriented, secure, and privacy-aware biometric systems.
Authors: David T. Hoffmann, Syed Haseeb Raza, Hanqiu Jiang, Denis Tananaev, Steffen Klingenhoefer, Martin Meinke
Abstract: Scene flow estimation is a foundational task for many robotic applications, including robust dynamic object detection, automatic labeling, and sensor synchronization. Two types of approaches to the problem have evolved: 1) Supervised and 2) optimization-based methods. Supervised methods are fast during inference and achieve high-quality results, however, they are limited by the need for large amounts of labeled training data and are susceptible to domain gaps. In contrast, unsupervised test-time optimization methods do not face the problem of domain gaps but usually suffer from substantial runtime, exhibit artifacts, or fail to converge to the right solution. In this work, we mitigate several limitations of existing optimization-based methods. To this end, we 1) introduce a simple voxel grid-based model that improves over the standard MLP-based formulation in multiple dimensions and 2) introduce a new multiframe loss formulation. 3) We combine both contributions in our new method, termed Floxels. On the Argoverse 2 benchmark, Floxels is surpassed only by EulerFlow among unsupervised methods while achieving comparable performance at a fraction of the computational cost. Floxels achieves a massive speedup of more than ~60 - 140x over EulerFlow, reducing the runtime from a day to 10 minutes per sequence. Over the faster but low-quality baseline, NSFP, Floxels achieves a speedup of ~14x.
Authors: Yue Gao, Hong-Xing Yu, Bo Zhu, Jiajun Wu
Abstract: We study reconstructing and predicting 3D fluid appearance and velocity from a single video. Current methods require multi-view videos for fluid reconstruction. We present FluidNexus, a novel framework that bridges video generation and physics simulation to tackle this task. Our key insight is to synthesize multiple novel-view videos as references for reconstruction. FluidNexus consists of two key components: (1) a novel-view video synthesizer that combines frame-wise view synthesis with video diffusion refinement for generating realistic videos, and (2) a physics-integrated particle representation coupling differentiable simulation and rendering to simultaneously facilitate 3D fluid reconstruction and prediction. To evaluate our approach, we collect two new real-world fluid datasets featuring textured backgrounds and object interactions. Our method enables dynamic novel view synthesis, future prediction, and interaction simulation from a single fluid video. Project website: https://yuegao.me/FluidNexus.
Authors: Zahra Gharaee, Scott C. Lowe, ZeMing Gong, Pablo Millan Arias, Nicholas Pellegrino, Austin T. Wang, Joakim Bruslund Haurum, Iuliia Zarubiieva, Lila Kari, Dirk Steinke, Graham W. Taylor, Paul Fieguth, Angel X. Chang
Abstract: As part of an ongoing worldwide effort to comprehend and monitor insect biodiversity, this paper presents the BIOSCAN-5M Insect dataset to the machine learning community and establish several benchmark tasks. BIOSCAN-5M is a comprehensive dataset containing multi-modal information for over 5 million insect specimens, and it significantly expands existing image-based biological datasets by including taxonomic labels, raw nucleotide barcode sequences, assigned barcode index numbers, geographical, and size information. We propose three benchmark experiments to demonstrate the impact of the multi-modal data types on the classification and clustering accuracy. First, we pretrain a masked language model on the DNA barcode sequences of the BIOSCAN-5M dataset, and demonstrate the impact of using this large reference library on species- and genus-level classification performance. Second, we propose a zero-shot transfer learning task applied to images and DNA barcodes to cluster feature embeddings obtained from self-supervised learning, to investigate whether meaningful clusters can be derived from these representation embeddings. Third, we benchmark multi-modality by performing contrastive learning on DNA barcodes, image data, and taxonomic information. This yields a general shared embedding space enabling taxonomic classification using multiple types of information and modalities. The code repository of the BIOSCAN-5M Insect dataset is available at https://github.com/bioscan-ml/BIOSCAN-5M.
Authors: Arthur Zhang, Harshit Sikchi, Amy Zhang, Joydeep Biswas
Abstract: We address the long-horizon mapless navigation problem: enabling robots to traverse novel environments without relying on high-definition maps or precise waypoints that specify exactly where to navigate. Achieving this requires overcoming two major challenges -- learning robust, generalizable perceptual representations of the environment without pre-enumerating all possible navigation factors and forms of perceptual aliasing and utilizing these learned representations to plan human-aligned navigation paths. Existing solutions struggle to generalize due to their reliance on hand-curated object lists that overlook unforeseen factors, end-to-end learning of navigation features from scarce large-scale robot datasets, and handcrafted reward functions that scale poorly to diverse scenarios. To overcome these limitations, we propose CREStE, the first method that learns representations and rewards for addressing the full mapless navigation problem without relying on large-scale robot datasets or manually curated features. CREStE leverages visual foundation models trained on internet-scale data to learn continuous bird's-eye-view representations capturing elevation, semantics, and instance-level features. To utilize learned representations for planning, we propose a counterfactual-based loss and active learning procedure that focuses on the most salient perceptual cues by querying humans for counterfactual trajectory annotations in challenging scenes. We evaluate CREStE in kilometer-scale navigation tasks across six distinct urban environments. CREStE significantly outperforms all state-of-the-art approaches with 70% fewer human interventions per mission, including a 2-kilometer mission in an unseen environment with just 1 intervention; showcasing its robustness and effectiveness for long-horizon mapless navigation. For videos and additional materials, see https://amrl.cs.utexas.edu/creste .
Authors: Zi Wang, Shiyi Lan, Xinglong Sun, Nadine Chang, Zhenxin Li, Zhiding Yu, Jose M. Alvarez
Abstract: Autonomous vehicle safety is crucial for the successful deployment of self-driving cars. However, most existing planning methods rely heavily on imitation learning, which limits their ability to leverage collision data effectively. Moreover, collecting collision or near-collision data is inherently challenging, as it involves risks and raises ethical and practical concerns. In this paper, we propose SafeFusion, a training framework to learn from collision data. Instead of over-relying on imitation learning, SafeFusion integrates safety-oriented metrics during training to enable collision avoidance learning. In addition, to address the scarcity of collision data, we propose CollisionGen, a scalable data generation pipeline to generate diverse, high-quality scenarios using natural language prompts, generative models, and rule-based filtering. Experimental results show that our approach improves planning performance in collision-prone scenarios by 56\% over previous state-of-the-art planners while maintaining effectiveness in regular driving situations. Our work provides a scalable and effective solution for advancing the safety of autonomous driving systems.
Authors: Yifei Gao, Jun Huang, Lei Wang, Ruiting Dai, Jun Cheng
Abstract: The emergence of 3D Gaussian Splatting (3D-GS) has significantly advanced 3D reconstruction by providing high fidelity and fast training speeds across various scenarios. While recent efforts have mainly focused on improving model structures to compress data volume or reduce artifacts during zoom-in and zoom-out operations, they often overlook an underlying issue: training sampling deficiency. In zoomed-in views, Gaussian primitives can appear unregulated and distorted due to their dilation limitations and the insufficient availability of scale-specific training samples. Consequently, incorporating pseudo-details that ensure the completeness and alignment of the scene becomes essential. In this paper, we introduce a new training method that integrates diffusion models and multi-scale training using pseudo-ground-truth data. This approach not only notably mitigates the dilation and zoomed-in artifacts but also enriches reconstructed scenes with precise details out of existing scenarios. Our method achieves state-of-the-art performance across various benchmarks and extends the capabilities of 3D reconstruction beyond training datasets.
Authors: Tao Zhong, Christine Allen-Blanchette
Abstract: We propose GAGrasp, a novel framework for dexterous grasp generation that leverages geometric algebra representations to enforce equivariance to SE(3) transformations. By encoding the SE(3) symmetry constraint directly into the architecture, our method improves data and parameter efficiency while enabling robust grasp generation across diverse object poses. Additionally, we incorporate a differentiable physics-informed refinement layer, which ensures that generated grasps are physically plausible and stable. Extensive experiments demonstrate the model's superior performance in generalization, stability, and adaptability compared to existing methods. Additional details at https://gagrasp.github.io/
Authors: Joshua Bird, Jan Blumenkamp, Amanda Prorok
Abstract: Cooperative Simultaneous Localization and Mapping (C-SLAM) enables multiple agents to work together in mapping unknown environments while simultaneously estimating their own positions. This approach enhances robustness, scalability, and accuracy by sharing information between agents, reducing drift, and enabling collective exploration of larger areas. In this paper, we present Decentralized Visual Monocular SLAM (DVM-SLAM), the first open-source decentralized monocular C-SLAM system. By only utilizing low-cost and light-weight monocular vision sensors, our system is well suited for small robots and micro aerial vehicles (MAVs). DVM-SLAM's real-world applicability is validated on physical robots with a custom collision avoidance framework, showcasing its potential in real-time multi-agent autonomous navigation scenarios. We also demonstrate comparable accuracy to state-of-the-art centralized monocular C-SLAM systems. We open-source our code and provide supplementary material online.
Authors: Senming Tan, Zhenyu Hou, Zhihao Zhang, Long Xu, Mengke Zhang, Zhaoqi He, Chao Xu, Fei Gao, Yanjun Cao
Abstract: Terrain analysis is critical for the practical application of ground mobile robots in real-world tasks, especially in outdoor unstructured environments. In this paper, we propose a novel spatial-temporal traversability assessment method, which aims to enable autonomous robots to effectively navigate through complex terrains. Our approach utilizes sparse Gaussian processes (SGP) to extract geometric features (curvature, gradient, elevation, etc.) directly from point cloud scans. These features are then used to construct a high-resolution local traversability map. Then, we design a spatial-temporal Bayesian Gaussian kernel (BGK) inference method to dynamically evaluate traversability scores, integrating historical and real-time data while considering factors such as slope, flatness, gradient, and uncertainty metrics. GPU acceleration is applied in the feature extraction step, and the system achieves real-time performance. Extensive simulation experiments across diverse terrain scenarios demonstrate that our method outperforms SOTA approaches in both accuracy and computational efficiency. Additionally, we develop an autonomous navigation framework integrated with the traversability map and validate it with a differential driven vehicle in complex outdoor environments. Our code will be open-source for further research and development by the community, https://github.com/ZJU-FAST-Lab/FSGP_BGK.
Authors: Yingfei Sun, Xu Gu, Wei Ji, Hanbin Zhao, Hao Fei, Yifang Yin, Roger Zimmermann
Abstract: Many studies combine text and audio to capture multi-modal information but they overlook the model's generalization ability on new datasets. Introducing new datasets may affect the feature space of the original dataset, leading to catastrophic forgetting. Meanwhile, large model parameters can significantly impact training performance. To address these limitations, we introduce a novel task called Text-Audio Incremental Learning (TAIL) task for text-audio retrieval, and propose a new method, PTAT, Prompt Tuning for Audio-Text incremental learning. This method utilizes prompt tuning to optimize the model parameters while incorporating an audio-text similarity and feature distillation module to effectively mitigate catastrophic forgetting. We benchmark our method and previous incremental learning methods on AudioCaps, Clotho, BBC Sound Effects and Audioset datasets, and our method outperforms previous methods significantly, particularly demonstrating stronger resistance to forgetting on older datasets. Compared to the full-parameters Finetune (Sequential) method, our model only requires 2.42\% of its parameters, achieving 4.46\% higher performance.
Authors: Luk\'a\v{s} Gajdo\v{s}ech, Hassan Ali, Jan-Gerrit Habekost, Martin Madaras, Matthias Kerzel, Stefan Wermter
Abstract: Datasets for object detection often do not account for enough variety of glasses, due to their transparent and reflective properties. Specifically, open-vocabulary object detectors, widely used in embodied robotic agents, fail to distinguish subclasses of glasses. This scientific gap poses an issue to robotic applications that suffer from accumulating errors between detection, planning, and action execution. The paper introduces a novel method for the acquisition of real-world data from RGB-D sensors that minimizes human effort. We propose an auto-labeling pipeline that generates labels for all the acquired frames based on the depth measurements. We provide a novel real-world glass object dataset that was collected on the Neuro-Inspired COLlaborator (NICOL), a humanoid robot platform. The data set consists of 7850 images recorded from five different cameras. We show that our trained baseline model outperforms state-of-the-art open-vocabulary approaches. In addition, we deploy our baseline model in an embodied agent approach to the NICOL platform, on which it achieves a success rate of 81% in a human-robot bartending scenario.
Authors: Cecilia Diana-Albelda, Roberto Alcover-Couso, \'Alvaro Garc\'ia-Mart\'in, Jesus Bescos, Marcos Escudero-Vi\~nolo
Abstract: Gliomas are brain tumours that stand out for their highly lethal and aggressive nature, which demands a precise approach in their diagnosis. Medical image segmentation plays a crucial role in the evaluation and follow-up of these tumours, allowing specialists to analyse their morphology. However, existing methods for automatic glioma segmentation often lack generalization capability across other brain tumour domains, require extensive computational resources, or fail to fully utilize the multi-parametric MRI (mp-MRI) data used to delineate them. In this work, we introduce GBT-SAM, a novel Generalizable Brain Tumour (GBT) framework that extends the Segment Anything Model (SAM) to brain tumour segmentation tasks. Our method employs a two-step training protocol: first, fine-tuning the patch embedding layer to process the entire mp-MRI modalities, and second, incorporating parameter-efficient LoRA blocks and a Depth-Condition block into the Vision Transformer (ViT) to capture inter-slice correlations. GBT-SAM achieves state-of-the-art performance on the Adult Glioma dataset (Dice Score of $93.54$) while demonstrating robust generalization across Meningioma, Pediatric Glioma, and Sub-Saharan Glioma datasets. Furthermore, GBT-SAM uses less than 6.5M trainable parameters, thus offering an efficient solution for brain tumour segmentation. \\ Our code and models are available at https://github.com/vpulab/med-sam-brain .
Authors: Maxime Burchi, Radu Timofte
Abstract: The DreamerV3 algorithm recently obtained remarkable performance across diverse environment domains by learning an accurate world model based on Recurrent Neural Networks (RNNs). Following the success of model-based reinforcement learning algorithms and the rapid adoption of the Transformer architecture for its superior training efficiency and favorable scaling properties, recent works such as STORM have proposed replacing RNN-based world models with Transformer-based world models using masked self-attention. However, despite the improved training efficiency of these methods, their impact on performance remains limited compared to the Dreamer algorithm, struggling to learn competitive Transformer-based world models. In this work, we show that the next state prediction objective adopted in previous approaches is insufficient to fully exploit the representation capabilities of Transformers. We propose to extend world model predictions to longer time horizons by introducing TWISTER (Transformer-based World model wIth contraSTivE Representations), a world model using action-conditioned Contrastive Predictive Coding to learn high-level temporal feature representations and improve the agent performance. TWISTER achieves a human-normalized mean score of 162% on the Atari 100k benchmark, setting a new record among state-of-the-art methods that do not employ look-ahead search.
Authors: Rohit Menon, Nils Dengler, Sicong Pan, Gokul Krishna Chenchani, Maren Bennewitz
Abstract: For scene understanding in unstructured environments, an accurate and uncertainty-aware metric-semantic mapping is required to enable informed action selection by autonomous systems.Existing mapping methods often suffer from overconfident semantic predictions, and sparse and noisy depth sensing, leading to inconsistent map representations. In this paper, we therefore introduce EvidMTL, a multi-task learning framework that uses evidential heads for depth estimation and semantic segmentation, enabling uncertainty-aware inference from monocular RGB images. To enable uncertainty-calibrated evidential multi-task learning, we propose a novel evidential depth loss function that jointly optimizes the belief strength of the depth prediction in conjunction with evidential segmentation loss. Building on this, we present EvidKimera, an uncertainty-aware semantic surface mapping framework, which uses evidential depth and semantics prediction for improved 3D metric-semantic consistency. We train and evaluate EvidMTL on the NYUDepthV2 and assess its zero-shot performance on ScanNetV2, demonstrating superior uncertainty estimation compared to conventional approaches while maintaining comparable depth estimation and semantic segmentation. In zero-shot mapping tests on ScanNetV2, EvidKimera outperforms Kimera in semantic surface mapping accuracy and consistency, highlighting the benefits of uncertainty-aware mapping and underscoring its potential for real-world robotic applications.
Authors: Adrian Chang, Kai Wang, Yuanbo Li, Manolis Savva, Angel X. Chang, Daniel Ritchie
Abstract: Data driven and autoregressive indoor scene synthesis systems generate indoor scenes automatically by suggesting and then placing objects one at a time. Empirical observations show that current systems tend to produce incomplete next object location distributions. We introduce a system which addresses this problem. We design a Domain Specific Language (DSL) that specifies functional constraints. Programs from our language take as input a partial scene and object to place. Upon execution they predict possible object placements. We design a generative model which writes these programs automatically. Available 3D scene datasets do not contain programs to train on, so we build upon previous work in unsupervised program induction to introduce a new program bootstrapping algorithm. In order to quantify our empirical observations we introduce a new evaluation procedure which captures how well a system models per-object location distributions. We ask human annotators to label all the possible places an object can go in a scene and show that our system produces per-object location distributions more consistent with human annotators. Our system also generates indoor scenes of comparable quality to previous systems and while previous systems degrade in performance when training data is sparse, our system does not degrade to the same degree.
Authors: Alessandro Scherl, Stefan Thalhammer, Bernhard Neuberger, Wilfried W\"ober, Jos\'e Grac\'ia-Rodr\'iguez
Abstract: Visual servoing enables robots to precisely position their end-effector relative to a target object. While classical methods rely on hand-crafted features and thus are universally applicable without task-specific training, they often struggle with occlusions and environmental variations, whereas learning-based approaches improve robustness but typically require extensive training. We present a visual servoing approach that leverages pretrained vision transformers for semantic feature extraction, combining the advantages of both paradigms while also being able to generalize beyond the provided sample. Our approach achieves full convergence in unperturbed scenarios and surpasses classical image-based visual servoing by up to 31.2\% relative improvement in perturbed scenarios. Even the convergence rates of learning-based methods are matched despite requiring no task- or object-specific training. Real-world evaluations confirm robust performance in end-effector positioning, industrial box manipulation, and grasping of unseen objects using only a reference from the same category. Our code and simulation environment are available at: https://alessandroscherl.github.io/ViT-VS/
Authors: Artin Saberpour Abadian, Yi-Chi Liao, Ata Otaran, Rishabh Dabral, Marie Muehlhaus, Christian Theobalt, Martin Schmitz, J\"urgen Steimle
Abstract: Supernumerary robotic limbs (SRLs) are robotic structures integrated closely with the user's body, which augment human physical capabilities and necessitate seamless, naturalistic human-machine interaction. For effective assistance in physical tasks, enabling SRLs to hand over objects to humans is crucial. Yet, designing heuristic-based policies for robots is time-consuming, difficult to generalize across tasks, and results in less human-like motion. When trained with proper datasets, generative models are powerful alternatives for creating naturalistic handover motions. We introduce 3HANDS, a novel dataset of object handover interactions between a participant performing a daily activity and another participant enacting a hip-mounted SRL in a naturalistic manner. 3HANDS captures the unique characteristics of SRL interactions: operating in intimate personal space with asymmetric object origins, implicit motion synchronization, and the user's engagement in a primary task during the handover. To demonstrate the effectiveness of our dataset, we present three models: one that generates naturalistic handover trajectories, another that determines the appropriate handover endpoints, and a third that predicts the moment to initiate a handover. In a user study (N=10), we compare the handover interaction performed with our method compared to a baseline. The findings show that our method was perceived as significantly more natural, less physically demanding, and more comfortable.
Authors: Hong Liu, Haosen Yang, Federica Eduati, Josien P. W. Pluim, Mitko Veta
Abstract: Leveraging multimodal data, particularly the integration of whole-slide histology images (WSIs) and transcriptomic profiles, holds great promise for improving cancer survival prediction. However, excessive redundancy in multimodal data can degrade model performance. In this paper, we propose Adaptive Prototype Learning (APL), a novel and effective approach for multimodal cancer survival analysis. APL adaptively learns representative prototypes in a data-driven manner, reducing redundancy while preserving critical information. Our method employs two sets of learnable query vectors that serve as a bridge between high-dimensional representations and survival prediction, capturing task-relevant features. Additionally, we introduce a multimodal mixed self-attention mechanism to enable cross-modal interactions, further enhancing information fusion. Extensive experiments on five benchmark cancer datasets demonstrate the superiority of our approach over existing methods. The code is available at https://github.com/HongLiuuuuu/APL.
Authors: Blaine Quackenbush, Paul J. Atzberger
Abstract: We introduce methods for obtaining pretrained Geometric Neural Operators (GNPs) that can serve as basal foundation models for use in obtaining geometric features. These can be used within data processing pipelines for machine learning tasks and numerical methods. We show how our GNPs can be trained to learn robust latent representations for the differential geometry of point-clouds to provide estimates of metric, curvature, and other shape-related features. We demonstrate how our pre-trained GNPs can be used (i) to estimate the geometric properties of surfaces of arbitrary shape and topologies with robustness in the presence of noise, (ii) to approximate solutions of geometric partial differential equations (PDEs) on manifolds, and (iii) to solve equations for shape deformations such as curvature driven flows. We also release a package of the codes and weights for using our pre-trained GNPs for processing point cloud representations. This allows for incorporating our pre-trained GNPs as components for reuse within existing and new data processing pipelines. The GNPs also can be used as part of numerical solvers involving geometry or as part of methods for performing inference and other geometric tasks.
Authors: Amirhossein Askari Farsangi, Ali Sharifi-Zarchi, Mohammad Hossein Rohban
Abstract: Acute Lymphoblastic Leukemia (ALL) is one of the most common types of childhood blood cancer. The quick start of the treatment process is critical to saving the patient's life, and for this reason, early diagnosis of this disease is essential. Examining the blood smear images of these patients is one of the methods used by expert doctors to diagnose this disease. Deep learning-based methods have numerous applications in medical fields, as they have significantly advanced in recent years. ALL diagnosis is not an exception in this field, and several machine learning-based methods for this problem have been proposed. In previous methods, high diagnostic accuracy was reported, but our work showed that this alone is not sufficient, as it can lead to models taking shortcuts and not making meaningful decisions. This issue arises due to the small size of medical training datasets. To address this, we constrained our model to follow a pipeline inspired by experts' work. We also demonstrated that, since a judgement based on only one image is insufficient, redefining the problem as a multiple-instance learning problem is necessary for achieving a practical result. Our model is the first to provide a solution to this problem in a multiple-instance learning setup. We introduced a novel pipeline for diagnosing ALL that approximates the process used by hematologists, is sensitive to disease biomarkers, and achieves an accuracy of 96.15%, an F1-score of 94.24%, a sensitivity of 97.56%, and a specificity of 90.91% on ALL IDB 1. Our method was further evaluated on an out-of-distribution dataset, which posed a challenging test and had acceptable performance. Notably, our model was trained on a relatively small dataset, highlighting the potential for our approach to be applied to other medical datasets with limited data availability.
Authors: Hongsin Lee, Seungju Cho, Changick Kim
Abstract: Adversarial training significantly improves adversarial robustness, but superior performance is primarily attained with large models. This substantial performance gap for smaller models has spurred active research into adversarial distillation (AD) to mitigate the difference. Existing AD methods leverage the teacher's logits as a guide. In contrast to these approaches, we aim to transfer another piece of knowledge from the teacher, the input gradient. In this paper, we propose a distillation module termed Indirect Gradient Distillation Module (IGDM) that indirectly matches the student's input gradient with that of the teacher. Experimental results show that IGDM seamlessly integrates with existing AD methods, significantly enhancing their performance. Particularly, utilizing IGDM on the CIFAR-100 dataset improves the AutoAttack accuracy from 28.06% to 30.32% with the ResNet-18 architecture and from 26.18% to 29.32% with the MobileNetV2 architecture when integrated into the SOTA method without additional data augmentation.
Authors: Xiang Xu, Lingdong Kong, Hui Shuai, Qingshan Liu
Abstract: LiDAR segmentation has become a crucial component of advanced autonomous driving systems. Recent range-view LiDAR segmentation approaches show promise for real-time processing. However, they inevitably suffer from corrupted contextual information and rely heavily on post-processing techniques for prediction refinement. In this work, we propose FRNet, a simple yet powerful method aimed at restoring the contextual information of range image pixels using corresponding frustum LiDAR points. First, a frustum feature encoder module is used to extract per-point features within the frustum region, which preserves scene consistency and is critical for point-level predictions. Next, a frustum-point fusion module is introduced to update per-point features hierarchically, enabling each point to extract more surrounding information through the frustum features. Finally, a head fusion module is used to fuse features at different levels for final semantic predictions. Extensive experiments conducted on four popular LiDAR segmentation benchmarks under various task setups demonstrate the superiority of FRNet. Notably, FRNet achieves 73.3% and 82.5% mIoU scores on the testing sets of SemanticKITTI and nuScenes. While achieving competitive performance, FRNet operates 5 times faster than state-of-the-art approaches. Such high efficiency opens up new possibilities for more scalable LiDAR segmentation. The code has been made publicly available at https://github.com/Xiangxu-0103/FRNet.
Authors: Zongxin Yang, Guikun Chen, Xiaodi Li, Wenguan Wang, Yi Yang
Abstract: Recent LLM-driven visual agents mainly focus on solving image-based tasks, which limits their ability to understand dynamic scenes, making it far from real-life applications like guiding students in laboratory experiments and identifying their mistakes. Hence, this paper explores DoraemonGPT, a comprehensive and conceptually elegant system driven by LLMs to understand dynamic scenes. Considering the video modality better reflects the ever-changing nature of real-world scenarios, we exemplify DoraemonGPT as a video agent. Given a video with a question/task, DoraemonGPT begins by converting the input video into a symbolic memory that stores task-related attributes. This structured representation allows for spatial-temporal querying and reasoning by well-designed sub-task tools, resulting in concise intermediate results. Recognizing that LLMs have limited internal knowledge when it comes to specialized domains (e.g., analyzing the scientific principles underlying experiments), we incorporate plug-and-play tools to assess external knowledge and address tasks across different domains. Moreover, a novel LLM-driven planner based on Monte Carlo Tree Search is introduced to explore the large planning space for scheduling various tools. The planner iteratively finds feasible solutions by backpropagating the result's reward, and multiple solutions can be summarized into an improved final answer. We extensively evaluate DoraemonGPT's effectiveness on three benchmarks and several in-the-wild scenarios. The code will be released at https://github.com/z-x-yang/DoraemonGPT.
Authors: Michelle R. Greene, Mariam Josyula, Wentao Si, Jennifer A. Hart
Abstract: Computer-based scene understanding has influenced fields ranging from urban planning to autonomous vehicle performance, yet little is known about how well these technologies work across social differences. We investigate the biases of deep convolutional neural networks (dCNNs) in scene classification, using nearly one million images from global and US sources, including user-submitted home photographs and Airbnb listings. We applied statistical models to quantify the impact of socioeconomic indicators such as family income, Human Development Index (HDI), and demographic factors from public data sources (CIA and US Census) on dCNN performance. Our analyses revealed significant socioeconomic bias, where pretrained dCNNs demonstrated lower classification accuracy, lower classification confidence, and a higher tendency to assign labels that could be offensive when applied to homes (e.g., "ruin", "slum"), especially in images from homes with lower socioeconomic status (SES). This trend is consistent across two datasets of international images and within the diverse economic and racial landscapes of the United States. This research contributes to understanding biases in computer vision, emphasizing the need for more inclusive and representative training datasets. By mitigating the bias in the computer vision pipelines, we can ensure fairer and more equitable outcomes for applied computer vision, including home valuation and smart home security systems. There is urgency in addressing these biases, which can significantly impact critical decisions in urban development and resource allocation. Our findings also motivate the development of AI systems that better understand and serve diverse communities, moving towards technology that equitably benefits all sectors of society.
Authors: Paul Gavrikov, Jovita Lukasik, Steffen Jung, Robert Geirhos, M. Jehanzeb Mirza, Margret Keuper, Janis Keuper
Abstract: Unlike traditional vision-only models, vision language models (VLMs) offer an intuitive way to access visual content through language prompting by combining a large language model (LLM) with a vision encoder. However, both the LLM and the vision encoder come with their own set of biases, cue preferences, and shortcuts, which have been rigorously studied in uni-modal models. A timely question is how such (potentially misaligned) biases and cue preferences behave under multi-modal fusion in VLMs. As a first step towards a better understanding, we investigate a particularly well-studied vision-only bias - the texture vs. shape bias and the dominance of local over global information. As expected, we find that VLMs inherit this bias to some extent from their vision encoders. Surprisingly, the multi-modality alone proves to have important effects on the model behavior, i.e., the joint training and the language querying change the way visual cues are processed. While this direct impact of language-informed training on a model's visual perception is intriguing, it raises further questions on our ability to actively steer a model's output so that its prediction is based on particular visual cues of the user's choice. Interestingly, VLMs have an inherent tendency to recognize objects based on shape information, which is different from what a plain vision encoder would do. Further active steering towards shape-based classifications through language prompts is however limited. In contrast, active VLM steering towards texture-based decisions through simple natural language prompts is often more successful. URL: https://github.com/paulgavrikov/vlm_shapebias
Authors: Yifei Huang, Guo Chen, Jilan Xu, Mingfang Zhang, Lijin Yang, Baoqi Pei, Hongjie Zhang, Lu Dong, Yali Wang, Limin Wang, Yu Qiao
Abstract: Being able to map the activities of others into one's own point of view is one fundamental human skill even from a very early age. Taking a step toward understanding this human ability, we introduce EgoExoLearn, a large-scale dataset that emulates the human demonstration following process, in which individuals record egocentric videos as they execute tasks guided by demonstration videos. Focusing on the potential applications in daily assistance and professional support, EgoExoLearn contains egocentric and demonstration video data spanning 120 hours captured in daily life scenarios and specialized laboratories. Along with the videos we record high-quality gaze data and provide detailed multimodal annotations, formulating a playground for modeling the human ability to bridge asynchronous procedural actions from different viewpoints. To this end, we present benchmarks such as cross-view association, cross-view action planning, and cross-view referenced skill assessment, along with detailed analysis. We expect EgoExoLearn can serve as an important resource for bridging the actions across views, thus paving the way for creating AI agents capable of seamlessly learning by observing humans in the real world. Code and data can be found at: https://github.com/OpenGVLab/EgoExoLearn
Authors: Emil Benedykciuk, Marcin Denkowski, Grzegorz W\'ojcik
Abstract: This paper introduces a novel approach to enhance the performance of pre-trained neural networks in medical image segmentation using gradient-based Neural Architecture Search (NAS) methods. We present the concept of Implantable Adaptive Cell (IAC), small modules identified through Partially-Connected DARTS based approach, designed to be injected into the skip connections of an existing and already trained U-shaped model. Unlike traditional NAS methods, our approach refines existing architectures without full retraining. Experiments on four medical datasets with MRI and CT images show consistent accuracy improvements on various U-Net configurations, with segmentation accuracy gain by approximately 5 percentage points across all validation datasets, with improvements reaching up to 11\%pt in the best-performing cases. The findings of this study not only offer a cost-effective alternative to the complete overhaul of complex models for performance upgrades but also indicate the potential applicability of our method to other architectures and problem domains.
Authors: Hu Wang, Salma Hassan, Yuyuan Liu, Congbo Ma, Yuanhong Chen, Yutong Xie, Mostafa Salem, Yu Tian, Jodie Avery, Louise Hull, Ian Reid, Mohammad Yaqub, Gustavo Carneiro
Abstract: In multi-modal learning, some modalities are more influential than others, and their absence can have a significant impact on classification/segmentation accuracy. Addressing this challenge, we propose a novel approach called Meta-learned Modality-weighted Knowledge Distillation (MetaKD), which enables multi-modal models to maintain high accuracy even when key modalities are missing. MetaKD adaptively estimates the importance weight of each modality through a meta-learning process. These learned importance weights guide a pairwise modality-weighted knowledge distillation process, allowing high-importance modalities to transfer knowledge to lower-importance ones, resulting in robust performance despite missing inputs. Unlike previous methods in the field, which are often task-specific and require significant modifications, our approach is designed to work in multiple tasks (e.g., segmentation and classification) with minimal adaptation. Experimental results on five prevalent datasets, including three Brain Tumor Segmentation datasets (BraTS2018, BraTS2019 and BraTS2020), the Alzheimer's Disease Neuroimaging Initiative (ADNI) classification dataset and the Audiovision-MNIST classification dataset, demonstrate the proposed model is able to outperform the compared models by a large margin. The code is available at https://github.com/billhhh/MetaKD.
Authors: Xinyi Wang, Grazziela Figueredo, Ruizhe Li, Wei Emma Zhang, Weitong Chen, Xin Chen
Abstract: Automatic radiology report generation can alleviate the workload for physicians and minimize regional disparities in medical resources, therefore becoming an important topic in the medical image analysis field. It is a challenging task, as the computational model needs to mimic physicians to obtain information from multi-modal input data (i.e., medical images, clinical information, medical knowledge, etc.), and produce comprehensive and accurate reports. Recently, numerous works have emerged to address this issue using deep-learning-based methods, such as transformers, contrastive learning, and knowledge-base construction. This survey summarizes the key techniques developed in the most recent works and proposes a general workflow for deep-learning-based report generation with five main components, including multi-modality data acquisition, data preparation, feature learning, feature fusion and interaction, and report generation. The state-of-the-art methods for each of these components are highlighted. Additionally, we summarize the latest developments in large model-based methods and model explainability, along with public datasets, evaluation methods, current challenges, and future directions in this field. We have also conducted a quantitative comparison between different methods in the same experimental setting. This is the most up-to-date survey that focuses on multi-modality inputs and data fusion for radiology report generation. The aim is to provide comprehensive and rich information for researchers interested in automatic clinical report generation and medical image analysis, especially when using multimodal inputs, and to assist them in developing new algorithms to advance the field.
Authors: Yibin Wang, Weizhong Zhang, Honghui Xu, Cheng Jin
Abstract: Scene text synthesis involves rendering specified texts onto arbitrary images. Current methods typically formulate this task in an end-to-end manner but lack effective character-level guidance during training. Besides, their text encoders, pre-trained on a single font type, struggle to adapt to the diverse font styles encountered in practical applications. Consequently, these methods suffer from character distortion, repetition, and absence, particularly in polystylistic scenarios. To this end, this paper proposes DreamText for high-fidelity scene text synthesis. Our key idea is to reconstruct the diffusion training process, introducing more refined guidance tailored to this task, to expose and rectify the model's attention at the character level and strengthen its learning of text regions. This transformation poses a hybrid optimization challenge, involving both discrete and continuous variables. To effectively tackle this challenge, we employ a heuristic alternate optimization strategy. Meanwhile, we jointly train the text encoder and generator to comprehensively learn and utilize the diverse font present in the training dataset. This joint training is seamlessly integrated into the alternate optimization process, fostering a synergistic relationship between learning character embedding and re-estimating character attention. Specifically, in each step, we first encode potential character-generated position information from cross-attention maps into latent character masks. These masks are then utilized to update the representation of specific characters in the current step, which, in turn, enables the generator to correct the character's attention in the subsequent steps. Both qualitative and quantitative results demonstrate the superiority of our method to the state of the art.
Authors: Xinyi Shang, Peng Sun, Tao Lin
Abstract: Recent advancements in dataset distillation have demonstrated the significant benefits of employing soft labels generated by pre-trained teacher models. In this paper, we introduce a novel perspective by emphasizing the full utilization of labels. We first conduct a comprehensive comparison of various loss functions for soft label utilization in dataset distillation, revealing that the model trained on the synthetic dataset exhibits high sensitivity to the choice of loss function for soft label utilization. This finding highlights the necessity of a universal loss function for training models on synthetic datasets. Building on these insights, we introduce an extremely simple yet surprisingly effective plug-and-play approach, GIFT, which encompasses soft label refinement and a cosine similarity-based loss function to efficiently leverage full label information. Extensive experiments indicate that GIFT consistently enhances state-of-the-art dataset distillation methods across various dataset scales, without incurring additional computational costs. Importantly, GIFT significantly enhances cross-optimizer generalization, an area previously overlooked. For instance, on ImageNet-1K with IPC = 10, GIFT enhances the state-of-the-art method RDED by 30.8% in cross-optimizer generalization. Our code is available at https://github.com/LINs-lab/GIFT.
Authors: Sreyan Ghosh, Chandra Kiran Reddy Evuru, Sonal Kumar, Utkarsh Tyagi, Oriol Nieto, Zeyu Jin, Dinesh Manocha
Abstract: Large Vision-Language Models (LVLMs) often produce responses that misalign with factual information, a phenomenon known as hallucinations. While hallucinations are well-studied, the exact causes behind them remain underexplored. In this paper, we first investigate the root causes of hallucinations in LVLMs. Our findings reveal that existing mitigation techniques primarily reduce hallucinations for visual recognition prompts-those that require simple descriptions of visual elements-but fail for cognitive prompts that demand deliberate reasoning. We identify the core issue as a lack of true visual perception in LVLMs: although they can accurately recognize visual elements, they struggle to fully interpret these elements in the context of the input prompt and effectively link this recognition to their internal knowledge, which is critical for reasoning. To address this gap, we introduce Visual Description Grounded Decoding (VDGD), a simple, robust, and training-free method designed to enhance visual perception and improve reasoning capabilities in LVLMs. VDGD works by first generating a detailed description of the image and appending it as a prefix to the instruction. During response generation, tokens are sampled based on their KL divergence to the description, favoring candidates with lower divergence. Experimental results on multiple visual reasoning benchmarks and LVLMs demonstrate that VDGD consistently outperforms existing baselines 2% - 33%. Finally, we introduce VaLLu, a benchmark designed for comprehensive evaluation of the cognitive capabilities of LVLMs.
Authors: Federico Magistri, Thomas L\"abe, Elias Marks, Sumanth Nagulavancha, Yue Pan, Claus Smitt, Lasse Klingbeil, Michael Halstead, Heiner Kuhlmann, Chris McCool, Jens Behley, Cyrill Stachniss
Abstract: As the world population is expected to reach 10 billion by 2050, our agricultural production system needs to double its productivity despite a decline of human workforce in the agricultural sector. Autonomous robotic systems are one promising pathway to increase productivity by taking over labor-intensive manual tasks like fruit picking. To be effective, such systems need to monitor and interact with plants and fruits precisely, which is challenging due to the cluttered nature of agricultural environments causing, for example, strong occlusions. Thus, being able to estimate the complete 3D shapes of objects in presence of occlusions is crucial for automating operations such as fruit harvesting. In this paper, we propose the first publicly available 3D shape completion dataset for agricultural vision systems. We provide an RGB-D dataset for estimating the 3D shape of fruits. Specifically, our dataset contains RGB-D frames of single sweet peppers in lab conditions but also in a commercial greenhouse. For each fruit, we additionally collected high-precision point clouds that we use as ground truth. For acquiring the ground truth shape, we developed a measuring process that allows us to record data of real sweet pepper plants, both in the lab and in the greenhouse with high precision, and determine the shape of the sensed fruits. We release our dataset, consisting of almost 7,000 RGB-D frames belonging to more than 100 different fruits. We provide segmented RGB-D frames, with camera intrinsics to easily obtain colored point clouds, together with the corresponding high-precision, occlusion-free point clouds obtained with a high-precision laser scanner. We additionally enable evaluation of shape completion approaches on a hidden test set through a public challenge on a benchmark server.
Authors: Roberto Di Via, Francesca Odone, Vito Paolo Pastore
Abstract: Deep neural networks have been extensively applied in the medical domain for various tasks, including image classification, segmentation, and landmark detection. However, their application is often hindered by data scarcity, both in terms of available annotations and images. This study introduces a novel application of denoising diffusion probabilistic models (DDPMs) to the landmark detection task, specifically addressing the challenge of limited annotated data in x-ray imaging. Our key innovation lies in leveraging DDPMs for self-supervised pre-training in landmark detection, a previously unexplored approach in this domain. This method enables accurate landmark detection with minimal annotated training data (as few as 50 images), surpassing both ImageNet supervised pre-training and traditional self-supervised techniques across three popular x-ray benchmark datasets. To our knowledge, this work represents the first application of diffusion models for self-supervised learning in landmark detection, which may offer a valuable pre-training approach in few-shot regimes, for mitigating data scarcity.
Authors: JunYong Choi, SeokYeong Lee, Haesol Park, Seung-Won Jung, Ig-Jae Kim, Junghyun Cho
Abstract: In this paper, we propose a scene-level inverse rendering framework that uses multi-view images to decompose the scene into geometry, SVBRDF, and 3D spatially-varying lighting. While multi-view images have been widely used for object-level inverse rendering, scene-level inverse rendering has primarily been studied using single-view images due to the lack of a dataset containing high dynamic range multi-view images with ground-truth geometry, material, and spatially-varying lighting. To improve the quality of scene-level inverse rendering, a novel framework called Multi-view Attention Inverse Rendering (MAIR) was recently introduced. MAIR performs scene-level multi-view inverse rendering by expanding the OpenRooms dataset, designing efficient pipelines to handle multi-view images, and splitting spatially-varying lighting. Although MAIR showed impressive results, its lighting representation is fixed to spherical Gaussians, which limits its ability to render images realistically. Consequently, MAIR cannot be directly used in applications such as material editing. Moreover, its multi-view aggregation networks have difficulties extracting rich features because they only focus on the mean and variance between multi-view features. In this paper, we propose its extended version, called MAIR++. MAIR++ addresses the aforementioned limitations by introducing an implicit lighting representation that accurately captures the lighting conditions of an image while facilitating realistic rendering. Furthermore, we design a directional attention-based multi-view aggregation network to infer more intricate relationships between views. Experimental results show that MAIR++ not only achieves better performance than MAIR and single-view-based methods, but also displays robust performance on unseen real-world scenes.
Authors: Jiancheng Pan, Yanxing Liu, Yuqian Fu, Muyuan Ma, Jiahao Li, Danda Pani Paudel, Luc Van Gool, Xiaomeng Huang
Abstract: Object detection, particularly open-vocabulary object detection, plays a crucial role in Earth sciences, such as environmental monitoring, natural disaster assessment, and land-use planning. However, existing open-vocabulary detectors, primarily trained on natural-world images, struggle to generalize to remote sensing images due to a significant data domain gap. Thus, this paper aims to advance the development of open-vocabulary object detection in remote sensing community. To achieve this, we first reformulate the task as Locate Anything on Earth (LAE) with the goal of detecting any novel concepts on Earth. We then developed the LAE-Label Engine which collects, auto-annotates, and unifies up to 10 remote sensing datasets creating the LAE-1M - the first large-scale remote sensing object detection dataset with broad category coverage. Using the LAE-1M, we further propose and train the novel LAE-DINO Model, the first open-vocabulary foundation object detector for the LAE task, featuring Dynamic Vocabulary Construction (DVC) and Visual-Guided Text Prompt Learning (VisGT) modules. DVC dynamically constructs vocabulary for each training batch, while VisGT maps visual features to semantic space, enhancing text features. We comprehensively conduct experiments on established remote sensing benchmark DIOR, DOTAv2.0, as well as our newly introduced 80-class LAE-80C benchmark. Results demonstrate the advantages of the LAE-1M dataset and the effectiveness of the LAE-DINO method.
Authors: Lucas M\"oller, Pascal Tilli, Ngoc Thang Vu, Sebastian Pad\'o
Abstract: Dual encoder architectures like CLIP models map two types of inputs into a shared embedding space and predict similarities between them. Despite their success, it is, however, not understood how these models compare their two inputs. Common first-order feature-attribution methods can only provide limited insights into dual-encoders since their predictions depend on feature-interactions rather than on individual features. In this paper, we first derive a second-order method enabling the attribution of predictions by any differentiable dual encoder onto feature-interactions between its inputs. Second, we apply our method to CLIP models and show that they learn fine-grained correspondences between parts of captions and regions in images. They match objects across input modes also account for mismatches. This visual-linguistic grounding ability, however, varies heavily between object classes and exhibits pronounced out-of-domain effects. We can identify individual errors as well as systematic failure categories including object coverage, unusual scenes and correlated contexts.
Authors: Bojian Li, Bo Liu, Xinning Yao, Jinghua Yue, Fugen Zhou
Abstract: Depth estimation is a cornerstone of 3D reconstruction and plays a vital role in minimally invasive endoscopic surgeries. However, most current depth estimation networks rely on traditional convolutional neural networks, which are limited in their ability to capture global information. Foundation models offer a promising approach to enhance depth estimation, but those models currently available are primarily trained on natural images, leading to suboptimal performance when applied to endoscopic images. In this work, we introduce a novel fine-tuning strategy for the Depth Anything Model and integrate it with an intrinsic-based unsupervised monocular depth estimation framework. Our approach includes a low-rank adaptation technique based on random vectors, which improves the model's adaptability to different scales. Additionally, we propose a residual block built on depthwise separable convolution to compensate for the transformer's limited ability to capture local features. Our experimental results on the SCARED dataset and Hamlyn dataset show that our method achieves state-of-the-art performance while minimizing the number of trainable parameters. Applying this method in minimally invasive endoscopic surgery can enhance surgeons' spatial awareness, thereby improving the precision and safety of the procedures.
Authors: Hossein Resani, Behrooz Nasihatkon, Mohammadreza Alimoradi Jazi
Abstract: We introduce a novel framework for Continual Learning in 3D object classification. Our approach, CL3D, is based on the selection of prototypes from each class using spectral clustering. For non-Euclidean data such as point clouds, spectral clustering can be employed as long as one can define a distance measure between pairs of samples. Choosing the appropriate distance measure enables us to leverage 3D geometric characteristics to identify representative prototypes for each class. We explore the effectiveness of clustering in the input space (3D points), local feature space (1024-dimensional points), and global feature space. We conduct experiments on the ModelNet40, ShapeNet, and ScanNet datasets, achieving state-of-the-art accuracy exclusively through the use of input space features. By leveraging the combined input, local, and global features, we have improved the state-of-the-art on ModelNet and ShapeNet, utilizing nearly half the memory used by competing approaches. For the challenging ScanNet dataset, our method enhances accuracy by 4.1% while consuming just 28% of the memory used by our competitors, demonstrating the scalability of our approach.
Authors: Kaijie Yin, Tian Gao, Hui Kong
Abstract: A prior global topological map (e.g., the OpenStreetMap, OSM) can boost the performance of autonomous mapping by a ground mobile robot. However, the prior map is usually incomplete due to lacking labeling in partial paths. To solve this problem, this paper proposes an OSM maker using airborne sensors carried by low-altitude aircraft, where the core of the OSM maker is a novel efficient pathfinder approach based on LiDAR and camera data, i.e., a binary dual-stream road segmentation model. Specifically, a multi-scale feature extraction based on the UNet architecture is implemented for images and point clouds. To reduce the effect caused by the sparsity of point cloud, an attention-guided gated block is designed to integrate image and point-cloud features. To optimize the model for edge deployment that significantly reduces storage footprint and computational demands, we propose a binarization streamline to each model component, including a variant of vision transformer (ViT) architecture as the encoder of the image branch, and new focal and perception losses to optimize the model training. The experimental results on two datasets demonstrate that our pathfinder method achieves SOTA accuracy with high efficiency in finding paths from the low-level airborne sensors, and we can create complete OSM prior maps based on the segmented road skeletons. Code and data are available at: \href{https://github.com/IMRL/Pathfinder}{https://github.com/IMRL/Pathfinder}.
URLs: https://github.com/IMRL/Pathfinder, https://github.com/IMRL/Pathfinder
Authors: {\L}ukasz Struski, Dawid Rymarczyk, Jacek Tabor
Abstract: In this work, we introduce InfoDisent, a hybrid approach to explainability based on the information bottleneck principle. InfoDisent enables the disentanglement of information in the final layer of any pretrained model into atomic concepts, which can be interpreted as prototypical parts. This approach merges the flexibility of post-hoc methods with the concept-level modeling capabilities of self-explainable neural networks, such as ProtoPNets. We demonstrate the effectiveness of InfoDisent through computational experiments and user studies across various datasets using modern backbones such as ViTs and convolutional networks. Notably, InfoDisent generalizes the prototypical parts approach to novel domains (ImageNet).
Authors: Amaia Cardiel, Eloi Zablocki, Elias Ramzi, Oriane Sim\'eoni, Matthieu Cord
Abstract: Vision Language Models (VLMs) have demonstrated remarkable capabilities in various open-vocabulary tasks, yet their zero-shot performance lags behind task-specific fine-tuned models, particularly in complex tasks like Referring Expression Comprehension (REC). Fine-tuning usually requires 'white-box' access to the model's architecture and weights, which is not always feasible due to proprietary or privacy concerns. In this work, we propose LLM-wrapper, a method for 'black-box' adaptation of VLMs for the REC task using Large Language Models (LLMs). LLM-wrapper capitalizes on the reasoning abilities of LLMs, improved with a light fine-tuning, to select the most relevant bounding box matching the referring expression, from candidates generated by a zero-shot black-box VLM. Our approach offers several advantages: it enables the adaptation of closed-source models without needing access to their internal workings, it is versatile as it works with any VLM, it transfers to new VLMs and datasets, and it allows for the adaptation of an ensemble of VLMs. We evaluate LLM-wrapper on multiple datasets using different VLMs and LLMs, demonstrating significant performance improvements and highlighting the versatility of our method. While LLM-wrapper is not meant to directly compete with standard white-box fine-tuning, it offers a practical and effective alternative for black-box VLM adaptation. Code and checkpoints are available at https://github.com/valeoai/LLM_wrapper .
Authors: Zhangshuo Qi, Junyi Ma, Jingyi Xu, Zijie Zhou, Luqi Cheng, Guangming Xiong
Abstract: Place recognition is a crucial component that enables autonomous vehicles to obtain localization results in GPS-denied environments. In recent years, multimodal place recognition methods have gained increasing attention. They overcome the weaknesses of unimodal sensor systems by leveraging complementary information from different modalities. However, most existing methods explore cross-modality correlations through feature-level or descriptor-level fusion, suffering from a lack of interpretability. Conversely, the recently proposed 3D Gaussian Splatting provides a new perspective on multimodal fusion by harmonizing different modalities into an explicit scene representation. In this paper, we propose a 3D Gaussian Splatting-based multimodal place recognition network dubbed GSPR. It explicitly combines multi-view RGB images and LiDAR point clouds into a spatio-temporally unified scene representation with the proposed Multimodal Gaussian Splatting. A network composed of 3D graph convolution and transformer is designed to extract spatio-temporal features and global descriptors from the Gaussian scenes for place recognition. Extensive evaluations on three datasets demonstrate that our method can effectively leverage complementary strengths of both multi-view cameras and LiDAR, achieving SOTA place recognition performance while maintaining solid generalization ability. Our open-source code will be released at https://github.com/QiZS-BIT/GSPR.
Authors: Feng Tian, Yixuan Li, Yichao Yan, Shanyan Guan, Yanhao Ge, Xiaokang Yang
Abstract: In the field of image editing, three core challenges persist: controllability, background preservation, and efficiency. Inversion-based methods rely on time-consuming optimization to preserve the features of the initial images, which results in low efficiency due to the requirement for extensive network inference. Conversely, inversion-free methods lack theoretical support for background similarity, as they circumvent the issue of maintaining initial features to achieve efficiency. As a consequence, none of these methods can achieve both high efficiency and background consistency. To tackle the challenges and the aforementioned disadvantages, we introduce PostEdit, a method that incorporates a posterior scheme to govern the diffusion sampling process. Specifically, a corresponding measurement term related to both the initial features and Langevin dynamics is introduced to optimize the estimated image generated by the given target prompt. Extensive experimental results indicate that the proposed PostEdit achieves state-of-the-art editing performance while accurately preserving unedited regions. Furthermore, the method is both inversion- and training-free, necessitating approximately 1.5 seconds and 18 GB of GPU memory to generate high-quality results.
Authors: Gao Yu Lee, Tanmoy Dam, Md Meftahul Ferdaus, Daniel Puiu Poenar, Vu Duong
Abstract: Image dehazing is crucial for clarifying images obscured by haze or fog, but current learning-based approaches is dependent on large volumes of training data and hence consumed significant computational power. Additionally, their performance is often inadequate under non-uniform or heavy haze. To address these challenges, we developed the Detail Recovery And Contrastive DehazeNet, which facilitates efficient and effective dehazing via a dense dilated inverted residual block and an attention-based detail recovery network that tailors enhancements to specific dehazed scene contexts. A major innovation is its ability to train effectively with limited data, achieved through a novel quadruplet loss-based contrastive dehazing paradigm. This approach distinctly separates hazy and clear image features while also distinguish lower-quality and higher-quality dehazed images obtained from each sub-modules of our network, thereby refining the dehazing process to a larger extent. Extensive tests on a variety of benchmarked haze datasets demonstrated the superiority of our approach. The code repository for this work is available at https://github.com/GreedYLearner1146/DRACO-DehazeNet.
Authors: Han Bao, Yue Huang, Yanbo Wang, Jiayi Ye, Xiangqi Wang, Xiuying Chen, Yue Zhao, Tianyi Zhou, Mohamed Elhoseiny, Xiangliang Zhang
Abstract: Large Vision-Language Models (LVLMs) have become essential for advancing the integration of visual and linguistic information. However, the evaluation of LVLMs presents significant challenges as the evaluation benchmark always demands lots of human cost for its construction, and remains static, lacking flexibility once constructed. Even though automatic evaluation has been explored in textual modality, the visual modality remains under-explored. As a result, in this work, we address a question: "Can LVLMs themselves be used to benchmark each other in the visual automatically domain?". We introduce AutoBench-V, an automated framework for serving evaluation on demand, i.e., benchmarking LVLMs based on specific aspects of model capability. AutoBench-V leverages text-to-image models to generate relevant image samples and then utilizes LVLMs to orchestrate visual question-answering (VQA) tasks, completing the evaluation process efficiently and flexibly. Through an extensive evaluation of nine popular LVLMs across five demanded user inputs (i.e., evaluation capabilities), the framework shows effectiveness and reliability.
Authors: Yichen He, Yuan Lin, Jianchao Wu, Hanchong Zhang, Yuchen Zhang, Ruicheng Le
Abstract: Existing large vision-language models (LVLMs) are largely limited to processing short, seconds-long videos and struggle with generating coherent descriptions for extended video spanning minutes or more. Long video description introduces new challenges, such as consistent character identification and plot-level descriptions incorporating both visual and audio information. To address these, we figure out audio-visual character identification, matching character names to each dialogue, as a key factor. We propose StoryTeller, a system for generating dense descriptions of long videos, incorporating both low-level visual concepts and high-level plot information. StoryTeller uses a multimodal large language model that integrates visual, audio, and text modalities to perform audio-visual character identification on minute-long video clips. The results are then fed into a LVLM to enhance consistency of video description. We validate our approach on movie description tasks and introduce MovieStory101, a dataset with dense descriptions for three-minute movie clips. To evaluate long video descriptions, we create StoryQA, a large set of multiple-choice questions for MovieStory101 test set. We assess descriptions by inputting them into GPT-4 to answer these questions, using accuracy as an automatic evaluation metric. Experiments show that StoryTeller outperforms all open and closed-source baselines on StoryQA, achieving 9.5% higher accuracy than the strongest baseline, Gemini-1.5-pro, and demonstrating a +15.56% advantage in human side-by-side evaluations. Additionally, incorporating audio-visual character identification from StoryTeller improves the performance of all video description models, with Gemini-1.5-pro and GPT-4o showing relative improvement of 5.5% and 13.0%, respectively, in accuracy on StoryQA.
Authors: Zhaoqing Wang, Xiaobo Xia, Runnan Chen, Dongdong Yu, Changhu Wang, Mingming Gong, Tongliang Liu
Abstract: This paper presents the Large Vision Diffusion Transformer (LaVin-DiT), a scalable and unified foundation model designed to tackle over 20 computer vision tasks in a generative framework. Unlike existing large vision models directly adapted from natural language processing architectures, which rely on less efficient autoregressive techniques and disrupt spatial relationships essential for vision data, LaVin-DiT introduces key innovations to optimize generative performance for vision tasks. First, to address the high dimensionality of visual data, we incorporate a spatial-temporal variational autoencoder that encodes data into a continuous latent space. Second, for generative modeling, we develop a joint diffusion transformer that progressively produces vision outputs. Third, for unified multi-task training, in-context learning is implemented. Input-target pairs serve as task context, which guides the diffusion transformer to align outputs with specific tasks within the latent space. During inference, a task-specific context set and test data as queries allow LaVin-DiT to generalize across tasks without fine-tuning. Trained on extensive vision datasets, the model is scaled from 0.1B to 3.4B parameters, demonstrating substantial scalability and state-of-the-art performance across diverse vision tasks. This work introduces a novel pathway for large vision foundation models, underscoring the promising potential of diffusion transformers. The code and models are available.
Authors: Bing Cao, Quanhao Lu, Jiekang Feng, Qilong Wang, Qinghua Hu, Pengfei Zhu
Abstract: The dynamic imbalance of the fore-background is a major challenge in video object counting, which is usually caused by the sparsity of target objects. This remains understudied in existing works and often leads to severe under-/over-prediction errors. To tackle this issue in video object counting, we propose a density-embedded Efficient Masked Autoencoder Counting (E-MAC) framework in this paper. To empower the model's representation ability on density regression, we develop a new $\mathtt{D}$ensity-$\mathtt{E}$mbedded $\mathtt{M}$asked m$\mathtt{O}$deling ($\mathtt{DEMO}$) method, which first takes the density map as an auxiliary modality to perform multimodal self-representation learning for image and density map. Although $\mathtt{DEMO}$ contributes to effective cross-modal regression guidance, it also brings in redundant background information, making it difficult to focus on the foreground regions. To handle this dilemma, we propose an efficient spatial adaptive masking derived from density maps to boost efficiency. Meanwhile, we employ an optical flow-based temporal collaborative fusion strategy to effectively capture the dynamic variations across frames, aligning features to derive multi-frame density residuals. The counting accuracy of the current frame is boosted by harnessing the information from adjacent frames. In addition, considering that most existing datasets are limited to human-centric scenarios, we first propose a large video bird counting dataset, DroneBird, in natural scenarios for migratory bird protection. Extensive experiments on three crowd datasets and our \textit{DroneBird} validate our superiority against the counterparts. The code and dataset are available.
Authors: Chenjie Cao, Chaohui Yu, Shang Liu, Fan Wang, Xiangyang Xue, Yanwei Fu
Abstract: We introduce MVGenMaster, a multi-view diffusion model enhanced with 3D priors to address versatile Novel View Synthesis (NVS) tasks. MVGenMaster leverages 3D priors that are warped using metric depth and camera poses, significantly enhancing both generalization and 3D consistency in NVS. Our model features a simple yet effective pipeline that can generate up to 100 novel views conditioned on variable reference views and camera poses with a single forward process. Additionally, we have developed a comprehensive large-scale multi-view image dataset called MvD-1M, comprising up to 1.6 million scenes, equipped with well-aligned metric depth to train MVGenMaster. Moreover, we present several training and model modifications to strengthen the model with scaled-up datasets. Extensive evaluations across in- and out-of-domain benchmarks demonstrate the effectiveness of our proposed method and data formulation. Models and codes will be released at https://github.com/ewrfcas/MVGenMaster/.
Authors: Zixuan Chen, Guangcong Wang, Jiahao Zhu, Jianhuang Lai, Xiaohua Xie
Abstract: 3D Gaussian Splatting (3DGS) has recently created impressive 3D assets for various applications. However, the copyright of these assets is not well protected as existing watermarking methods are not suited for the 3DGS rendering pipeline considering security, capacity, and invisibility. Besides, these methods often require hours or even days for optimization, limiting the application scenarios. In this paper, we propose GuardSplat, an innovative and efficient framework that effectively protects the copyright of 3DGS assets. Specifically, 1) We first propose a CLIP-guided Message Decoupling Optimization module for training the message decoder, leveraging CLIP's aligning capability and rich representations to achieve a high extraction accuracy with minimal optimization costs, presenting exceptional capacity and efficiency. 2) Then, we propose a Spherical-harmonic-aware (SH-aware) Message Embedding module tailored for 3DGS, which employs a set of SH offsets to seamlessly embed the message into the SH features of each 3D Gaussian while maintaining the original 3D structure. It enables the 3DGS assets to be watermarked with minimal fidelity trade-offs and also prevents malicious users from removing the messages from the model files, meeting the demands for invisibility and security. 3) We further propose an Anti-distortion Message Extraction module to improve robustness against various visual distortions. Extensive experiments demonstrate that GuardSplat outperforms state-of-the-art and achieves fast optimization speed. Project page: https://narcissusex.github.io/GuardSplat, and Code: https://github.com/NarcissusEx/GuardSplat.
URLs: https://narcissusex.github.io/GuardSplat,, https://github.com/NarcissusEx/GuardSplat.
Authors: Taesung Kwon, Jong Chul Ye
Abstract: In this paper, we propose a novel framework for solving high-definition video inverse problems using latent image diffusion models. Building on recent advancements in spatio-temporal optimization for video inverse problems using image diffusion models, our approach leverages latent-space diffusion models to achieve enhanced video quality and resolution. To address the high computational demands of processing high-resolution frames, we introduce a pseudo-batch consistent sampling strategy, allowing efficient operation on a single GPU. Additionally, to improve temporal consistency, we present pseudo-batch inversion, an initialization technique that incorporates informative latents from the measurement. By integrating with SDXL, our framework achieves state-of-the-art video reconstruction across a wide range of spatio-temporal inverse problems, including complex combinations of frame averaging and various spatial degradations, such as deblurring, super-resolution, and inpainting. Unlike previous methods, our approach supports multiple aspect ratios (landscape, vertical, and square) and delivers HD-resolution reconstructions (exceeding 1280x720) in under 6 seconds per frame on a single NVIDIA 4090 GPU.
Authors: Xuanyu Zhang, Zecheng Tang, Zhipei Xu, Runyi Li, Youmin Xu, Bin Chen, Feng Gao, Jian Zhang
Abstract: With the rapid growth of generative AI and its widespread application in image editing, new risks have emerged regarding the authenticity and integrity of digital content. Existing versatile watermarking approaches suffer from trade-offs between tamper localization precision and visual quality. Constrained by the limited flexibility of previous framework, their localized watermark must remain fixed across all images. Under AIGC-editing, their copyright extraction accuracy is also unsatisfactory. To address these challenges, we propose OmniGuard, a novel augmented versatile watermarking approach that integrates proactive embedding with passive, blind extraction for robust copyright protection and tamper localization. OmniGuard employs a hybrid forensic framework that enables flexible localization watermark selection and introduces a degradation-aware tamper extraction network for precise localization under challenging conditions. Additionally, a lightweight AIGC-editing simulation layer is designed to enhance robustness across global and local editing. Extensive experiments show that OmniGuard achieves superior fidelity, robustness, and flexibility. Compared to the recent state-of-the-art approach EditGuard, our method outperforms it by 4.25dB in PSNR of the container image, 20.7% in F1-Score under noisy conditions, and 14.8% in average bit accuracy.
Authors: Rui Chen, Zehuan Wu, Yichen Liu, Yuxin Guo, Jingcheng Ni, Haifeng Xia, Siyu Xia
Abstract: The creation of diverse and realistic driving scenarios has become essential to enhance perception and planning capabilities of the autonomous driving system. However, generating long-duration, surround-view consistent driving videos remains a significant challenge. To address this, we present UniMLVG, a unified framework designed to generate extended street multi-perspective videos under precise control. By integrating single- and multi-view driving videos into the training data, our approach updates a DiT-based diffusion model equipped with cross-frame and cross-view modules across three stages with multi training objectives, substantially boosting the diversity and quality of generated visual content. Importantly, we propose an innovative explicit viewpoint modeling approach for multi-view video generation to effectively improve motion transition consistency. Capable of handling various input reference formats (e.g., text, images, or video), our UniMLVG generates high-quality multi-view videos according to the corresponding condition constraints such as 3D bounding boxes or frame-level text descriptions. Compared to the best models with similar capabilities, our framework achieves improvements of 48.2% in FID and 35.2% in FVD.
Authors: Wenbo Huang, Jinghui Zhang, Guang Li, Lei Zhang, Shuoyuan Wang, Fang Dong, Jiahui Jin, Takahiro Ogawa, Miki Haseyama
Abstract: In few-shot action recognition (FSAR), long sub-sequences of video naturally express entire actions more effectively. However, the high computational complexity of mainstream Transformer-based methods limits their application. Recent Mamba demonstrates efficiency in modeling long sequences, but directly applying Mamba to FSAR overlooks the importance of local feature modeling and alignment. Moreover, long sub-sequences within the same class accumulate intra-class variance, which adversely impacts FSAR performance. To solve these challenges, we propose a Matryoshka MAmba and CoNtrasTive LeArning framework (Manta). Firstly, the Matryoshka Mamba introduces multiple Inner Modules to enhance local feature representation, rather than directly modeling global features. An Outer Module captures dependencies of timeline between these local features for implicit temporal alignment. Secondly, a hybrid contrastive learning paradigm, combining both supervised and unsupervised methods, is designed to mitigate the negative effects of intra-class variance accumulation. The Matryoshka Mamba and the hybrid contrastive learning paradigm operate in two parallel branches within Manta, enhancing Mamba for FSAR of long sub-sequence. Manta achieves new state-of-the-art performance on prominent benchmarks, including SSv2, Kinetics, UCF101, and HMDB51. Extensive empirical studies prove that Manta significantly improves FSAR of long sub-sequence from multiple perspectives.
Authors: Rui Qian, Xin Yin, Dejing Dou
Abstract: Current Large Multimodal Models (LMMs) empowered visual grounding typically rely on $\texttt{
Authors: Alexander Valverde, Luis Solano
Abstract: In Costa Rica, an average of 5 tons of seashells are extracted from ecosystems annually. Confiscated seashells, cannot be returned to their ecosystems due to the lack of origin recognition. To address this issue, we developed a convolutional neural network (CNN) specifically for seashell identification. We built a dataset from scratch, consisting of approximately 19000 images from the Pacific and Caribbean coasts. Using this dataset, the model achieved a classification accuracy exceeding 85%. The model has been integrated into a user-friendly application, which has classified over 36,000 seashells to date, delivering real-time results within 3 seconds per image. To further enhance the system's accuracy, an anomaly detection mechanism was incorporated to filter out irrelevant or anomalous inputs, ensuring only valid seashell images are processed.
Authors: Daniel Rossi, Guido Borghi, Roberto Vezzani
Abstract: Designing efficient neural networks for embedded devices is a critical challenge, particularly in applications requiring real-time performance, such as aerial imaging with drones and UAVs for emergency responses. In this work, we introduce TakuNet, a novel light-weight architecture which employs techniques such as depth-wise convolutions and an early downsampling stem to reduce computational complexity while maintaining high accuracy. It leverages dense connections for fast convergence during training and uses 16-bit floating-point precision for optimization on embedded hardware accelerators. Experimental evaluation on two public datasets shows that TakuNet achieves near-state-of-the-art accuracy in classifying aerial images of emergency situations, despite its minimal parameter count. Real-world tests on embedded devices, namely Jetson Orin Nano and Raspberry Pi, confirm TakuNet's efficiency, achieving more than 650 fps on the 15W Jetson board, making it suitable for real-time AI processing on resource-constrained platforms and advancing the applicability of drones in emergency scenarios. The code and implementation details are publicly released.
Authors: Xiangyu Gao, Yu Dai, Benliu Qiu, Lanxiao Wang, Heqian Qiu, Hongliang Li
Abstract: Owing to large-scale image-text contrastive training, pre-trained vision language model (VLM) like CLIP shows superior open-vocabulary recognition ability. Most existing open-vocabulary object detectors attempt to utilize the pre-trained VLMs to attain generalized representation. F-ViT uses the pre-trained visual encoder as the backbone network and freezes it during training. However, its frozen backbone doesn't benefit from the labeled data to strengthen the representation for detection. Therefore, we propose a novel two-branch backbone network, named as \textbf{V}iT-Feature-\textbf{M}odulated Multi-Scale \textbf{C}onvolutional Network (VMCNet), which consists of a trainable convolutional branch, a frozen pre-trained ViT branch and a VMC module. The trainable CNN branch could be optimized with labeled data while the frozen pre-trained ViT branch could keep the representation ability derived from large-scale pre-training. Then, the proposed VMC module could modulate the multi-scale CNN features with the representations from ViT branch. With this proposed mixed structure, the detector is more likely to discover objects of novel categories. Evaluated on two popular benchmarks, our method boosts the detection performance on novel category and outperforms state-of-the-art methods. On OV-COCO, the proposed method achieves 44.3 AP$_{50}^{\mathrm{novel}}$ with ViT-B/16 and 48.5 AP$_{50}^{\mathrm{novel}}$ with ViT-L/14. On OV-LVIS, VMCNet with ViT-B/16 and ViT-L/14 reaches 27.8 and 38.4 mAP$_{r}$.
Authors: Eun-Sol Park, MiSo Park, Seung Park, Yong-Goo Shin
Abstract: Transferability, the ability of adversarial examples crafted for one model to deceive other models, is crucial for black-box attacks. Despite advancements in attack methods for semantic segmentation, transferability remains limited, reducing their effectiveness in real-world applications. To address this, we introduce the Feature Similarity Projected Gradient Descent (FSPGD) attack, a novel black-box approach that enhances both attack performance and transferability. Unlike conventional segmentation attacks that rely on output predictions for gradient calculation, FSPGD computes gradients from intermediate layer features. Specifically, our method introduces a loss function that targets local information by comparing features between clean images and adversarial examples, while also disrupting contextual information by accounting for spatial relationships between objects. Experiments on Pascal VOC 2012 and Cityscapes datasets demonstrate that FSPGD achieves superior transferability and attack performance, establishing a new state-of-the-art benchmark. Code is available at https://github.com/KU-AIVS/FSPGD.
Authors: Zhifei Yang, Keyang Lu, Chao Zhang, Jiaxing Qi, Hanqi Jiang, Ruifei Ma, Shenglin Yin, Yifan Xu, Mingzhe Xing, Zhen Xiao, Jieyi Long, Xiangde Liu, Guangyao Zhai
Abstract: Controllable 3D scene generation has extensive applications in virtual reality and interior design, where the generated scenes should exhibit high levels of realism and controllability in terms of geometry. Scene graphs provide a suitable data representation that facilitates these applications. However, current graph-based methods for scene generation are constrained to text-based inputs and exhibit insufficient adaptability to flexible user inputs, hindering the ability to precisely control object geometry. To address this issue, we propose MMGDreamer, a dual-branch diffusion model for scene generation that incorporates a novel Mixed-Modality Graph, visual enhancement module, and relation predictor. The mixed-modality graph allows object nodes to integrate textual and visual modalities, with optional relationships between nodes. It enhances adaptability to flexible user inputs and enables meticulous control over the geometry of objects in the generated scenes. The visual enhancement module enriches the visual fidelity of text-only nodes by constructing visual representations using text embeddings. Furthermore, our relation predictor leverages node representations to infer absent relationships between nodes, resulting in more coherent scene layouts. Extensive experimental results demonstrate that MMGDreamer exhibits superior control of object geometry, achieving state-of-the-art scene generation performance. Project page: https://yangzhifeio.github.io/project/MMGDreamer.
Authors: Jonathan Roberts, Mohammad Reza Taesiri, Ansh Sharma, Akash Gupta, Samuel Roberts, Ioana Croitoru, Simion-Vlad Bogolin, Jialu Tang, Florian Langer, Vyas Raina, Vatsal Raina, Hanyi Xiong, Vishaal Udandarao, Jingyi Lu, Shiyang Chen, Sam Purkis, Tianshuo Yan, Wenye Lin, Gyungin Shin, Qiaochu Yang, Anh Totti Nguyen, David I. Atkinson, Aaditya Baranwal, Alexandru Coca, Mikah Dang, Sebastian Dziadzio, Jakob D. Kunz, Kaiqu Liang, Alexander Lo, Brian Pulfer, Steven Walton, Charig Yang, Kai Han, Samuel Albanie
Abstract: Large Multimodal Models (LMMs) exhibit major shortfalls when interpreting images and, by some measures, have poorer spatial cognition than small children or animals. Despite this, they attain high scores on many popular visual benchmarks, with headroom rapidly eroded by an ongoing surge of model progress. To address this, there is a pressing need for difficult benchmarks that remain relevant for longer. We take this idea to its limit by introducing ZeroBench-a lightweight visual reasoning benchmark that is entirely impossible for contemporary frontier LMMs. Our benchmark consists of 100 manually curated questions and 334 less difficult subquestions. We evaluate 20 LMMs on ZeroBench, all of which score 0.0%, and rigorously analyse the errors. To encourage progress in visual understanding, we publicly release ZeroBench.
Authors: Sujan Sai Gannamaneni, Rohil Prakash Rao, Michael Mock, Maram Akila, Stefan Wrobel
Abstract: Slice discovery methods (SDMs) are prominent algorithms for finding systematic weaknesses in DNNs. They identify top-k semantically coherent slices/subsets of data where a DNN-under-test has low performance. For being directly useful, slices should be aligned with human-understandable and relevant dimensions, which, for example, are defined by safety and domain experts as part of the operational design domain (ODD). While SDMs can be applied effectively on structured data, their application on image data is complicated by the lack of semantic metadata. To address these issues, we present an algorithm that combines foundation models for zero-shot image classification to generate semantic metadata with methods for combinatorial search to find systematic weaknesses in images. In contrast to existing approaches, ours identifies weak slices that are in line with pre-defined human-understandable dimensions. As the algorithm includes foundation models, its intermediate and final results may not always be exact. Therefore, we include an approach to address the impact of noisy metadata. We validate our algorithm on both synthetic and real-world datasets, demonstrating its ability to recover human-understandable systematic weaknesses. Furthermore, using our approach, we identify systematic weaknesses of multiple pre-trained and publicly available state-of-the-art computer vision DNNs.
Authors: Wei Dai, Jun Liu
Abstract: Efficient evaluation of three-dimensional (3D) medical images is crucial for diagnostic and therapeutic practices in healthcare. Recent years have seen a substantial uptake in applying deep learning and computer vision to analyse and interpret medical images. Traditional approaches, such as convolutional neural networks (CNNs) and vision transformers (ViTs), face significant computational challenges, prompting the need for architectural advancements. Recent efforts have led to the introduction of novel architectures like the ``Mamba'' model as alternative solutions to traditional CNNs or ViTs. The Mamba model excels in the linear processing of one-dimensional data with low computational demands. However, Mamba's potential for 3D medical image analysis remains underexplored and could face significant computational challenges as the dimension increases. This manuscript presents MobileViM, a streamlined architecture for efficient segmentation of 3D medical images. In the MobileViM network, we invent a new dimension-independent mechanism and a dual-direction traversing approach to incorporate with a vision-Mamba-based framework. MobileViM also features a cross-scale bridging technique to improve efficiency and accuracy across various medical imaging modalities. With these enhancements, MobileViM achieves segmentation speeds exceeding 90 frames per second (FPS) on a single graphics processing unit (i.e., NVIDIA RTX 4090). This performance is over 24 FPS faster than the state-of-the-art deep learning models for processing 3D images with the same computational resources. In addition, experimental evaluations demonstrate that MobileViM delivers superior performance, with Dice similarity scores reaching 92.72%, 86.69%, 80.46%, and 77.43% for PENGWIN, BraTS2024, ATLAS, and Toothfairy2 datasets, respectively, which significantly surpasses existing models.
Authors: Cuong V. Nguyen, Hieu X. Nguyen, Dung D. Pham Minh, Cuong D. Do
Abstract: Automated ECG diagnosis has seen significant advancements with deep learning techniques, but real-world applications still face challenges when dealing with scanned paper ECGs. In this study, we explore multi-label classification of ECGs extracted from scanned images, moving beyond traditional binary classification (normal/abnormal). We evaluate the performance of multiple deep neural network architectures, including AlexNet, VGG, ResNet, and Vision Transformer, on scanned ECG datasets. Our comparative analysis examines model accuracy, robustness to image artifacts, and generalizability across different ECG conditions. Additionally, we investigate whether ECG signals extracted from scanned images retain sufficient diagnostic information for reliable automated classification. The findings highlight the strengths and limitations of each architecture, providing insights into the feasibility of image-based ECG diagnosis and its potential integration into clinical workflows.
Authors: Thanh Trung Vu, Andreas Kofler, Kostas Papafitsoros
Abstract: We extend a recently introduced deep unrolling framework for learning spatially varying regularisation parameters in inverse imaging problems to the case of Total Generalised Variation (TGV). The framework combines a deep convolutional neural network (CNN) inferring the two spatially varying TGV parameters with an unrolled algorithmic scheme that solves the corresponding variational problem. The two subnetworks are jointly trained end-to-end in a supervised fashion and as such the CNN learns to compute those parameters that drive the reconstructed images as close to the ground truth as possible. Numerical results in image denoising and MRI reconstruction show a significant qualitative and quantitative improvement compared to the best TGV scalar parameter case as well as to other approaches employing spatially varying parameters computed by unsupervised methods. We also observe that the inferred spatially varying parameter maps have a consistent structure near the image edges, asking for further theoretical investigations. In particular, the parameter that weighs the first-order TGV term has a triple-edge structure with alternating high-low-high values whereas the one that weighs the second-order term attains small values in a large neighbourhood around the edges.
Authors: Bingxuan Li, Yiwei Wang, Jiuxiang Gu, Kai-Wei Chang, Nanyun Peng
Abstract: Chart generation aims to generate code to produce charts satisfying the desired visual properties, e.g., texts, layout, color, and type. It has great potential to empower the automatic professional report generation in financial analysis, research presentation, education, and healthcare. In this work, we build a vision-language model (VLM) based multi-agent framework for effective automatic chart generation. Generating high-quality charts requires both strong visual design skills and precise coding capabilities that embed the desired visual properties into code. Such a complex multi-modal reasoning process is difficult for direct prompting of VLMs. To resolve these challenges, we propose METAL, a multi-agent framework that decomposes the task of chart generation into the iterative collaboration among specialized agents. METAL achieves 5.2% improvement over the current best result in the chart generation task. The METAL framework exhibits the phenomenon of test-time scaling: its performance increases monotonically as the logarithmic computational budget grows from 512 to 8192 tokens. In addition, we find that separating different modalities during the critique process of METAL boosts the self-correction capability of VLMs in the multimodal context.
Authors: Zikuan Li, Qiaoyun Wu, Jialin Zhang, Kaijun Zhang, Jun Wang
Abstract: Spiking neural networks (SNNs), inspired by the spiking computation paradigm of the biological neural systems, have exhibited superior energy efficiency in 2D classification tasks over traditional artificial neural networks (ANNs). However, the regression potential of SNNs has not been well explored, especially in 3D point cloud processing. In this paper, we propose noise-injected spiking graph convolutional networks to leverage the full regression potential of SNNs in 3D point cloud denoising. Specifically, we first emulate the noise-injected neuronal dynamics to build noise-injected spiking neurons. On this basis, we design noise-injected spiking graph convolution for promoting disturbance-aware spiking representation learning on 3D points. Starting from the spiking graph convolution, we build two SNN-based denoising networks. One is a purely spiking graph convolutional network, which achieves low accuracy loss compared with some ANN-based alternatives, while resulting in significantly reduced energy consumption on two benchmark datasets, PU-Net and PC-Net. The other is a hybrid architecture that combines ANN-based learning with a high performance-efficiency trade-off in just a few time steps. Our work lights up SNN's potential for 3D point cloud denoising, injecting new perspectives of exploring the deployment on neuromorphic chips while paving the way for developing energy-efficient 3D data acquisition devices.
Authors: MohammadHossein Rezaei, Yicheng Fu, Phil Cuvin, Caleb Ziems, Yanzhe Zhang, Hao Zhu, Diyi Yang
Abstract: Human activity is moderated by norms. However, machines are often trained without explicit supervision on norm understanding and reasoning, especially when the norms are grounded in a physical and social context. To improve and evaluate the normative reasoning capability of vision-language models (VLMs), we present EgoNormia $\|\epsilon\|$, consisting of 1,853 ego-centric videos of human interactions, each of which has two related questions evaluating both the prediction and justification of normative actions. The normative actions encompass seven categories: safety, privacy, proxemics, politeness, cooperation, coordination/proactivity, and communication/legibility. To compile this dataset at scale, we propose a novel pipeline leveraging video sampling, automatic answer generation, filtering, and human validation. Our work demonstrates that current state-of-the-art vision-language models lack robust norm understanding, scoring a maximum of 45% on EgoNormia (versus a human bench of 92%). Our analysis of performance in each dimension highlights the significant risks of safety, privacy, and the lack of collaboration and communication capability when applied to real-world agents. We additionally show that through a retrieval-based generation method, it is possible to use EgoNormia to enhance normative reasoning in VLMs.
Authors: Yuqian Chen, Leo Zekelman, Yui Lo, Suheyla Cetin-Karayumak, Tengfei Xue, Yogesh Rathi, Nikos Makris, Fan Zhang, Weidong Cai, Lauren J. O'Donnell
Abstract: Tractography parcellation classifies streamlines reconstructed from diffusion MRI into anatomically defined fiber tracts for clinical and research applications. However, clinical scans often have incomplete fields of view (FOV) where brain regions are partially imaged, leading to partial or truncated fiber tracts. To address this challenge, we introduce TractCloud-FOV, a deep learning framework that robustly parcellates tractography under conditions of incomplete FOV. We propose a novel training strategy, FOV-Cut Augmentation (FOV-CA), in which we synthetically cut tractograms to simulate a spectrum of real-world inferior FOV cutoff scenarios. This data augmentation approach enriches the training set with realistic truncated streamlines, enabling the model to achieve superior generalization. We evaluate the proposed TractCloud-FOV on both synthetically cut tractography and two real-life datasets with incomplete FOV. TractCloud-FOV significantly outperforms several state-of-the-art methods on all testing datasets in terms of streamline classification accuracy, generalization ability, tract anatomical depiction, and computational efficiency. Overall, TractCloud-FOV achieves efficient and consistent tractography parcellation in diffusion MRI with incomplete FOV.
Authors: Xiwen Liang, Min Lin, Weiqi Ruan, Rongtao Xu, Yuecheng Liu, Jiaqi Chen, Bingqian Lin, Yuzheng Zhuang, Xiaodan Liang
Abstract: Existing methods for vision-language task planning excel in short-horizon tasks but often fall short in complex, long-horizon planning within dynamic environments. These challenges primarily arise from the difficulty of effectively training models to produce high-quality reasoning processes for long-horizon tasks. To address this, we propose Structured Preference Optimization (SPO), which aims to enhance reasoning and action selection in long-horizon task planning through structured preference evaluation and optimized training strategies. Specifically, SPO introduces: 1) Preference-Based Scoring and Optimization, which systematically evaluates reasoning chains based on task relevance, visual grounding, and historical consistency; and 2) Curriculum-Guided Training, where the model progressively adapts from simple to complex tasks, improving its generalization ability in long-horizon scenarios and enhancing reasoning robustness. To advance research in vision-language long-horizon task planning, we introduce ExtendaBench, a comprehensive benchmark covering 1,509 tasks across VirtualHome and Habitat 2.0, categorized into ultra-short, short, medium, and long tasks. Experimental results demonstrate that SPO significantly improves reasoning quality and final decision accuracy, outperforming prior methods on long-horizon tasks and underscoring the effectiveness of preference-driven optimization in vision-language task planning. Specifically, SPO achieves a +5.98% GCR and +4.68% SR improvement in VirtualHome and a +3.30% GCR and +2.11% SR improvement in Habitat over the best-performing baselines.
Authors: Benedikt Blumenstiel, Nassim Ait Ali Braham, Conrad M Albrecht, Stefano Maurogiovanni, Paolo Fraccaro
Abstract: This technical report presents SSL4EO-S12 v1.1, a multimodal, multitemporal Earth Observation dataset designed for pretraining large-scale foundation models. Building on the success of SSL4EO-S12 v1.0, the new version addresses the previous challenges of data misalignment and a limited data structure for low-barrier, analysis-ready EO processing. SSL4EO-S12 v1.1 covers the world's 10,000 largest cities and its surroundings within a 50 km radius across four seasons, resulting in a diverse collection of nearly one million patches. SSL4EO-S12 v1.1 packages the data in Zarr file format for cloud-efficient loading and representation of meta-information such as including cloud masks and geolocation. Released under the CC-BY-4.0 license, SSL4EO-S12 v1.1 facilitates open research and provides a robust foundation for future advancements in self-supervised learning and geospatial analysis. The dataset is available online through https://datapub.fz-juelich.de/ssl4eo-s12, and we provided additional resources at https://github.com/DLR-MF-DAS/SSL4EO-S12-v1.1.
URLs: https://datapub.fz-juelich.de/ssl4eo-s12,, https://github.com/DLR-MF-DAS/SSL4EO-S12-v1.1.
Authors: Wenke E, Chao Yuan, Li Li, Yixin Sun, Yona Falinie A. Gaus, Amir Atapour-Abarghouei, Toby P. Breckon
Abstract: We present Dur360BEV, a novel spherical camera autonomous driving dataset equipped with a high-resolution 128-channel 3D LiDAR and a RTK-refined GNSS/INS system, along with a benchmark architecture designed to generate Bird-Eye-View (BEV) maps using only a single spherical camera. This dataset and benchmark address the challenges of BEV generation in autonomous driving, particularly by reducing hardware complexity through the use of a single 360-degree camera instead of multiple perspective cameras. Within our benchmark architecture, we propose a novel spherical-image-to-BEV module that leverages spherical imagery and a refined sampling strategy to project features from 2D to 3D. Our approach also includes an innovative application of focal loss, specifically adapted to address the extreme class imbalance often encountered in BEV segmentation tasks, that demonstrates improved segmentation performance on the Dur360BEV dataset. The results show that our benchmark not only simplifies the sensor setup but also achieves competitive performance.
Authors: Wenhui Lei, Anqi Li, Yusheng Tan, Hanyu Chen, Xiaofan Zhang
Abstract: Foundation Models (FMs) in computational pathology (CPath) have significantly advanced the extraction of meaningful features from histopathology image datasets, achieving strong performance across various clinical tasks. Despite their impressive performance, these models often exhibit variability when applied to different tasks, prompting the need for a unified framework capable of consistently excelling across various applications. In this work, we propose Shazam, a novel framework designed to efficiently combine multiple CPath models. Unlike previous approaches that train a fixed-parameter FM, Shazam dynamically extracts and refines information from diverse FMs for each specific task. To ensure that each FM contributes effectively without dominance, a novel distillation strategy is applied, guiding the student model with features from all teacher models, which enhances its generalization ability. Experimental results on two pathology patch classification datasets demonstrate that Shazam outperforms existing CPath models and other fusion methods. Its lightweight, flexible design makes it a promising solution for improving CPath analysis in real-world settings. Code will be available at https://github.com/Tuner12/Shazam.
Authors: Ziyang Zhang, Yang Yu, Yucheng Chen, Xulei Yang, Si Yong Yeo
Abstract: Despite significant progress in Vision-Language Pre-training (VLP), current approaches predominantly emphasize feature extraction and cross-modal comprehension, with limited attention to generating or transforming visual content. This gap hinders the model's ability to synthesize coherent and novel visual representations from textual prompts, thereby reducing the effectiveness of multi-modal learning. In this work, we propose MedUnifier, a unified VLP framework tailored for medical data. MedUnifier seamlessly integrates text-grounded image generation capabilities with multi-modal learning strategies, including image-text contrastive alignment, image-text matching and image-grounded text generation. Unlike traditional methods that reply on continuous visual representations, our approach employs visual vector quantization, which not only facilitates a more cohesive learning strategy for cross-modal understanding but also enhances multi-modal generation quality by effectively leveraging discrete representations. Our framework's effectiveness is evidenced by the experiments on established benchmarks, including uni-modal tasks (supervised fine-tuning), cross-modal tasks (image-text retrieval and zero-shot image classification), and multi-modal tasks (medical report generation, image synthesis), where it achieves state-of-the-art performance across various tasks. MedUnifier also offers a highly adaptable tool for a wide range of language and vision tasks in healthcare, marking advancement toward the development of a generalizable AI model for medical applications.
Authors: Sijin Sun, Ming Deng, Xingrui Yu, Xinyu Xi, Liangbin Zhao
Abstract: Metal defect detection is critical in industrial quality assurance, yet existing methods struggle with grayscale variations and complex defect states, limiting its robustness. To address these challenges, this paper proposes a Self-Adaptive Gamma Context-Aware SSM-based model(GCM-DET). This advanced detection framework integrating a Dynamic Gamma Correction (GC) module to enhance grayscale representation and optimize feature extraction for precise defect reconstruction. A State-Space Search Management (SSM) architecture captures robust multi-scale features, effectively handling defects of varying shapes and scales. Focal Loss is employed to mitigate class imbalance and refine detection accuracy. Additionally, the CD5-DET dataset is introduced, specifically designed for port container maintenance, featuring significant grayscale variations and intricate defect patterns. Experimental results demonstrate that the proposed model achieves substantial improvements, with mAP@0.5 gains of 27.6\%, 6.6\%, and 2.6\% on the CD5-DET, NEU-DET, and GC10-DET datasets.
Authors: Tian Gao, Zhiyuan Zhang, Yu Zhang, Huajun Liu, Kaijie Yin, Chengzhong Xu, Hui Kong
Abstract: Model binarization has made significant progress in enabling real-time and energy-efficient computation for convolutional neural networks (CNN), offering a potential solution to the deployment challenges faced by Vision Transformers (ViTs) on edge devices. However, due to the structural differences between CNN and Transformer architectures, simply applying binary CNN strategies to the ViT models will lead to a significant performance drop. To tackle this challenge, we propose BHViT, a binarization-friendly hybrid ViT architecture and its full binarization model with the guidance of three important observations. Initially, BHViT utilizes the local information interaction and hierarchical feature aggregation technique from coarse to fine levels to address redundant computations stemming from excessive tokens. Then, a novel module based on shift operations is proposed to enhance the performance of the binary Multilayer Perceptron (MLP) module without significantly increasing computational overhead. In addition, an innovative attention matrix binarization method based on quantization decomposition is proposed to evaluate the token's importance in the binarized attention matrix. Finally, we propose a regularization loss to address the inadequate optimization caused by the incompatibility between the weight oscillation in the binary layers and the Adam Optimizer. Extensive experimental results demonstrate that our proposed algorithm achieves SOTA performance among binary ViT methods.
Authors: Isaac Aguirre, Ivan Sipiran, Gabriel Monta\~nana
Abstract: In this paper, we explore a self-supervised model that learns to detect the symmetry of a single object without requiring a dataset-relying solely on the input object itself. We hypothesize that the symmetry of an object can be determined by its intrinsic features, eliminating the need for large datasets during training. Additionally, we design a self-supervised learning strategy that removes the necessity of ground truth labels. These two key elements make our approach both effective and efficient, addressing the prohibitive costs associated with constructing large, labeled datasets for this task. The novelty of our method lies in computing features for each point on the object based on the idea that symmetric points should exhibit similar visual appearances. To achieve this, we leverage features extracted from a foundational image model to compute a visual descriptor for the points. This approach equips the point cloud with visual features that facilitate the optimization of our self-supervised model. Experimental results demonstrate that our method surpasses the state-of-the-art models trained on large datasets. Furthermore, our model is more efficient, effective, and operates with minimal computational and data resources.
Authors: Nuria Alina Chandra, Ryan Murtfeldt, Lin Qiu, Arnab Karmakar, Hannah Lee, Emmanuel Tanumihardja, Kevin Farhat, Ben Caffee, Sejin Paik, Changyeon Lee, Jongwook Choi, Aerin Kim, Oren Etzioni
Abstract: In the age of increasingly realistic generative AI, robust deepfake detection is essential for mitigating fraud and disinformation. While many deepfake detectors report high accuracy on academic datasets, we show that these academic benchmarks are out of date and not representative of real-world deepfakes. We introduce Deepfake-Eval-2024, a new deepfake detection benchmark consisting of in-the-wild deepfakes collected from social media and deepfake detection platform users in 2024. Deepfake-Eval-2024 consists of 45 hours of videos, 56.5 hours of audio, and 1,975 images, encompassing the latest manipulation technologies. The benchmark contains diverse media content from 88 different websites in 52 different languages. We find that the performance of open-source state-of-the-art deepfake detection models drops precipitously when evaluated on Deepfake-Eval-2024, with AUC decreasing by 50\% for video, 48\% for audio, and 45\% for image models compared to previous benchmarks. We also evaluate commercial deepfake detection models and models finetuned on Deepfake-Eval-2024, and find that they have superior performance to off-the-shelf open-source models, but do not yet reach the accuracy of deepfake forensic analysts. The dataset is available at https://github.com/nuriachandra/Deepfake-Eval-2024.
Authors: Wenqi Guo, Yiyang Du, Shan Du
Abstract: Gas leakage poses a significant hazard that requires prevention. Traditionally, human inspection has been used for detection, a slow and labour-intensive process. Recent research has applied machine learning techniques to this problem, yet there remains a shortage of high-quality, publicly available datasets. This paper introduces a synthetic dataset featuring diverse backgrounds, interfering foreground objects, diverse leak locations, and precise segmentation ground truth. We propose a zero-shot method that combines background subtraction, zero-shot object detection, filtering, and segmentation to leverage this dataset. Experimental results indicate that our approach significantly outperforms baseline methods based solely on background subtraction and zero-shot object detection with segmentation, reaching an IoU of 69\% overall. We also present an analysis of various prompt configurations and threshold settings to provide deeper insights into the performance of our method. The code and dataset will be released after publication.
Authors: Jingzhou Luo, Yang Liu, Weixing Chen, Zhen Li, Yaowei Wang, Guanbin Li, Liang Lin
Abstract: 3D Question Answering (3D QA) requires the model to comprehensively understand its situated 3D scene described by the text, then reason about its surrounding environment and answer a question under that situation. However, existing methods usually rely on global scene perception from pure 3D point clouds and overlook the importance of rich local texture details from multi-view images. Moreover, due to the inherent noise in camera poses and complex occlusions, there exists significant feature degradation and reduced feature robustness problems when aligning 3D point cloud with multi-view images. In this paper, we propose a Dual-vision Scene Perception Network (DSPNet), to comprehensively integrate multi-view and point cloud features to improve robustness in 3D QA. Our Text-guided Multi-view Fusion (TGMF) module prioritizes image views that closely match the semantic content of the text. To adaptively fuse back-projected multi-view images with point cloud features, we design the Adaptive Dual-vision Perception (ADVP) module, enhancing 3D scene comprehension. Additionally, our Multimodal Context-guided Reasoning (MCGR) module facilitates robust reasoning by integrating contextual information across visual and linguistic modalities. Experimental results on SQA3D and ScanQA datasets demonstrate the superiority of our DSPNet. Codes will be available at https://github.com/LZ-CH/DSPNet.
Authors: Zhumei Wang, Zechen Hu, Ruoxi Guo, Huaijin Pi, Ziyong Feng, Sida Peng, Xiaowei Zhou
Abstract: Recovering absolute poses in the world coordinate system from monocular views presents significant challenges. Two primary issues arise in this context. Firstly, existing methods rely on 3D motion data for training, which requires collection in limited environments. Acquiring such 3D labels for new actions in a timely manner is impractical, severely restricting the model's generalization capabilities. In contrast, 2D poses are far more accessible and easier to obtain. Secondly, estimating a person's absolute position in metric space from a single viewpoint is inherently more complex. To address these challenges, we introduce Mocap-2-to-3, a novel framework that decomposes intricate 3D motions into 2D poses, leveraging 2D data to enhance 3D motion reconstruction in diverse scenarios and accurately predict absolute positions in the world coordinate system. We initially pretrain a single-view diffusion model with extensive 2D data, followed by fine-tuning a multi-view diffusion model for view consistency using publicly available 3D data. This strategy facilitates the effective use of large-scale 2D data. Additionally, we propose an innovative human motion representation that decouples local actions from global movements and encodes geometric priors of the ground, ensuring the generative model learns accurate motion priors from 2D data. During inference, this allows for the gradual recovery of global movements, resulting in more plausible positioning. We evaluate our model's performance on real-world datasets, demonstrating superior accuracy in motion and absolute human positioning compared to state-of-the-art methods, along with enhanced generalization and scalability. Our code will be made publicly available.
Authors: Ping Chen, Xingpeng Zhang, Zhaoxiang Liu, Huan Hu, Xiang Liu, Kai Wang, Min Wang, Yanlin Qian, Shiguo Lian
Abstract: In this research, we propose a novel denoising diffusion model based on shortest-path modeling that optimizes residual propagation to enhance both denoising efficiency and quality. Drawing on Denoising Diffusion Implicit Models (DDIM) and insights from graph theory, our model, termed the Shortest Path Diffusion Model (ShortDF), treats the denoising process as a shortest-path problem aimed at minimizing reconstruction error. By optimizing the initial residuals, we improve the efficiency of the reverse diffusion process and the quality of the generated samples. Extensive experiments on multiple standard benchmarks demonstrate that ShortDF significantly reduces diffusion time (or steps) while enhancing the visual fidelity of generated samples compared to prior arts. This work, we suppose, paves the way for interactive diffusion-based applications and establishes a foundation for rapid data generation. Code is available at https://github.com/UnicomAI/ShortDF
Authors: Li Lun, Kunyu Feng, Qinglong Ni, Ling Liang, Yuan Wang, Ying Li, Dunshan Yu, Xiaoxin Cui
Abstract: Spiking neural networks (SNNs) have shown their competence in handling spatial-temporal event-based data with low energy consumption. Similar to conventional artificial neural networks (ANNs), SNNs are also vulnerable to gradient-based adversarial attacks, wherein gradients are calculated by spatial-temporal back-propagation (STBP) and surrogate gradients (SGs). However, the SGs may be invisible for an inference-only model as they do not influence the inference results, and current gradient-based attacks are ineffective for binary dynamic images captured by the dynamic vision sensor (DVS). While some approaches addressed the issue of invisible SGs through universal SGs, their SGs lack a correlation with the victim model, resulting in sub-optimal performance. Moreover, the imperceptibility of existing SNN-based binary attacks is still insufficient. In this paper, we introduce an innovative potential-dependent surrogate gradient (PDSG) method to establish a robust connection between the SG and the model, thereby enhancing the adaptability of adversarial attacks across various models with invisible SGs. Additionally, we propose the sparse dynamic attack (SDA) to effectively attack binary dynamic images. Utilizing a generation-reduction paradigm, SDA can fully optimize the sparsity of adversarial perturbations. Experimental results demonstrate that our PDSG and SDA outperform state-of-the-art SNN-based attacks across various models and datasets. Specifically, our PDSG achieves 100% attack success rate on ImageNet, and our SDA obtains 82% attack success rate by modifying only 0.24% of the pixels on CIFAR10DVS. The code is available at https://github.com/ryime/PDSG-SDA .
Authors: Khoi Anh Nguyen, Linh Yen Vu, Thang Dinh Duong, Thuan Nguyen Duong, Huy Thanh Nguyen, Vinh Quang Dinh
Abstract: Visual Question Answering (VQA) is a multimodal task requiring reasoning across textual and visual inputs, which becomes particularly challenging in low-resource languages like Vietnamese due to linguistic variability and the lack of high-quality datasets. Traditional methods often rely heavily on extensive annotated datasets, computationally expensive pipelines, and large pre-trained models, specifically in the domain of Vietnamese VQA, limiting their applicability in such scenarios. To address these limitations, we propose a training framework that combines a paraphrase-based feature augmentation module with a dynamic curriculum learning strategy. Explicitly, augmented samples are considered "easy" while raw samples are regarded as "hard". The framework then utilizes a mechanism that dynamically adjusts the ratio of easy to hard samples during training, progressively modifying the same dataset to increase its difficulty level. By enabling gradual adaptation to task complexity, this approach helps the Vietnamese VQA model generalize well, thus improving overall performance. Experimental results show consistent improvements on the OpenViVQA dataset and mixed outcomes on the ViVQA dataset, highlighting both the potential and challenges of our approach in advancing VQA for Vietnamese language.
Authors: Nimra Dilawar, Sara Nadeem, Javed Iqbal, Waqas Sultani, Mohsen Ali
Abstract: Existing generic unsupervised domain adaptation approaches require access to both a large labeled source dataset and a sufficient unlabeled target dataset during adaptation. However, collecting a large dataset, even if unlabeled, is a challenging and expensive endeavor, especially in medical imaging. In addition, constraints such as privacy issues can result in cases where source data is unavailable. Taking in consideration these challenges, we propose MIAdapt, an adaptive approach for Microscopic Imagery Adaptation as a solution for Source-free Few-shot Domain Adaptive Object detection (SF-FSDA). We also define two competitive baselines (1) Faster-FreeShot and (2) MT-FreeShot. Extensive experiments on the challenging M5-Malaria and Raabin-WBC datasets validate the effectiveness of MIAdapt. Without using any image from the source domain MIAdapt surpasses state-of-the-art source-free UDA (SF-UDA) methods by +21.3% mAP and few-shot domain adaptation (FSDA) approaches by +4.7% mAP on Raabin-WBC. Our code and models will be publicly available.
Authors: ChenTong Wang, Jincheng Gao, Fei Zhu, Abderrahim Halimi, C\'edric Richard
Abstract: Transformers have shown significant success in hyperspectral unmixing (HU). However, challenges remain. While multi-scale and long-range spatial correlations are essential in unmixing tasks, current Transformer-based unmixing networks, built on Vision Transformer (ViT) or Swin-Transformer, struggle to capture them effectively. Additionally, current Transformer-based unmixing networks rely on the linear mixing model, which lacks the flexibility to accommodate scenarios where nonlinear effects are significant. To address these limitations, we propose a multi-scale Dilated Transformer-based unmixing network for nonlinear HU (DTU-Net). The encoder employs two branches. The first one performs multi-scale spatial feature extraction using Multi-Scale Dilated Attention (MSDA) in the Dilated Transformer, which varies dilation rates across attention heads to capture long-range and multi-scale spatial correlations. The second one performs spectral feature extraction utilizing 3D-CNNs with channel attention. The outputs from both branches are then fused to integrate multi-scale spatial and spectral information, which is subsequently transformed to estimate the abundances. The decoder is designed to accommodate both linear and nonlinear mixing scenarios. Its interpretability is enhanced by explicitly modeling the relationships between endmembers, abundances, and nonlinear coefficients in accordance with the polynomial post-nonlinear mixing model (PPNMM). Experiments on synthetic and real datasets validate the effectiveness of the proposed DTU-Net compared to PPNMM-derived methods and several advanced unmixing networks.
Authors: Wenqiao Li, Yao Gu, Xintao Chen, Xiaohao Xu, Ming Hu, Xiaonan Huang, Yingna Wu
Abstract: Humans detect real-world object anomalies by perceiving, interacting, and reasoning based on object-conditioned physical knowledge. The long-term goal of Industrial Anomaly Detection (IAD) is to enable machines to autonomously replicate this skill. However, current IAD algorithms are largely developed and tested on static, semantically simple datasets, which diverge from real-world scenarios where physical understanding and reasoning are essential. To bridge this gap, we introduce the Physics Anomaly Detection (Phys-AD) dataset, the first large-scale, real-world, physics-grounded video dataset for industrial anomaly detection. Collected using a real robot arm and motor, Phys-AD provides a diverse set of dynamic, semantically rich scenarios. The dataset includes more than 6400 videos across 22 real-world object categories, interacting with robot arms and motors, and exhibits 47 types of anomalies. Anomaly detection in Phys-AD requires visual reasoning, combining both physical knowledge and video content to determine object abnormality. We benchmark state-of-the-art anomaly detection methods under three settings: unsupervised AD, weakly-supervised AD, and video-understanding AD, highlighting their limitations in handling physics-grounded anomalies. Additionally, we introduce the Physics Anomaly Explanation (PAEval) metric, designed to assess the ability of visual-language foundation models to not only detect anomalies but also provide accurate explanations for their underlying physical causes. Our dataset and benchmark will be publicly available.
Authors: Xiaojun Bi, Shuo Li, Ziyue Wang, Fuwen Luo, Weizheng Qiao, Lu Han, Ziwei Sun, Peng Li, Yang Liu
Abstract: Dongba pictographs are the only pictographs still in use in the world. They have pictorial ideographic features, and their symbols carry rich cultural and contextual information. Due to the lack of relevant datasets, existing research has difficulty in advancing the study of semantic understanding of Dongba pictographs. To this end, we propose DongbaMIE, the first multimodal dataset for semantic understanding and extraction of Dongba pictographs. The dataset consists of Dongba pictograph images and their corresponding Chinese semantic annotations. It contains 23,530 sentence-level and 2,539 paragraph-level images, covering four semantic dimensions: objects, actions, relations, and attributes. We systematically evaluate the GPT-4o, Gemini-2.0, and Qwen2-VL models. Experimental results show that the F1 scores of GPT-4o and Gemini in the best object extraction are only 3.16 and 3.11 respectively. The F1 score of Qwen2-VL after supervised fine-tuning is only 11.49. These results suggest that current large multimodal models still face significant challenges in accurately recognizing the diverse semantic information in Dongba pictographs. The dataset can be obtained from this URL.
Authors: Wei Li, Bing Hu, Rui Shao, Leyang Shen, Liqiang Nie
Abstract: First-person video assistants are highly anticipated to enhance our daily lives through online video dialogue. However, existing online video assistants often sacrifice assistant efficacy for real-time efficiency by processing low-frame-rate videos with coarse-grained visual features.To overcome the trade-off between efficacy and efficiency, we propose "Fast & Slow Video-Language Thinker" as an onLIne videO assistaNt, LION-FS, achieving real-time, proactive, temporally accurate, and contextually precise responses. LION-FS adopts a two-stage optimization strategy: 1)Fast Path: Routing-Based Response Determination evaluates frame-by-frame whether an immediate response is necessary. To enhance response determination accuracy and handle higher frame-rate inputs efficiently, we employ Token Aggregation Routing to dynamically fuse spatiotemporal features without increasing token numbers, while utilizing Token Dropping Routing to eliminate redundant features. 2)Slow Path: Multi-granularity Keyframe Augmentation optimizes keyframes during response generation. To provide comprehensive and detailed responses beyond atomic actions constrained by training data, fine-grained spatial features and human-environment interaction features are extracted through multi-granular pooling. These features are further integrated into a meticulously designed multimodal Thinking Template to guide more precise response generation. Comprehensive evaluations on online video tasks demonstrate that LION-FS achieves state-of-the-art efficacy and efficiency.
Authors: Rickard Br\"uel-Gabrielsson, Tongzhou Wang, Manel Baradad, Justin Solomon
Abstract: Despite dropout's ubiquity in machine learning, its effectiveness as a form of data augmentation remains under-explored. We address two key questions: (i) When is dropout effective as an augmentation strategy? (ii) Is dropout uniquely effective under these conditions? To explore these questions, we propose Deep Augmentation, a network- and modality-agnostic method that applies dropout or PCA transformations to targeted layers in neural networks. Through extensive experiments on contrastive learning tasks in NLP, computer vision, and graph learning, we find that uniformly applying dropout across layers does not consistently improve performance. Instead, dropout proves most beneficial in deeper layers and can be matched by alternative augmentations (e.g., PCA). We also show that a stop-gradient operation is critical for ensuring dropout functions effectively as an augmentation, and that performance trends invert when moving from contrastive tasks to supervised tasks. Our analysis suggests that Deep Augmentation helps mitigate inter-layer co-adaptation -- a notable issue in self-supervised learning due to the absence of labeled data. Drawing on these insights, we outline a procedure for selecting the optimal augmentation layer and demonstrate that Deep Augmentation can outperform traditional input-level augmentations. This simple yet powerful approach can be seamlessly integrated into a wide range of architectures and modalities, yielding notable gains in both performance and generalization.
Authors: Antonio Emanuele Cin\`a, Francesco Villani, Maura Pintor, Lea Sch\"onherr, Battista Biggio, Marcello Pelillo
Abstract: Evaluating the adversarial robustness of deep networks to gradient-based attacks is challenging. While most attacks consider $\ell_2$- and $\ell_\infty$-norm constraints to craft input perturbations, only a few investigate sparse $\ell_1$- and $\ell_0$-norm attacks. In particular, $\ell_0$-norm attacks remain the least studied due to the inherent complexity of optimizing over a non-convex and non-differentiable constraint. However, evaluating adversarial robustness under these attacks could reveal weaknesses otherwise left untested with more conventional $\ell_2$- and $\ell_\infty$-norm attacks. In this work, we propose a novel $\ell_0$-norm attack, called $\sigma$-zero, which leverages a differentiable approximation of the $\ell_0$ norm to facilitate gradient-based optimization, and an adaptive projection operator to dynamically adjust the trade-off between loss minimization and perturbation sparsity. Extensive evaluations using MNIST, CIFAR10, and ImageNet datasets, involving robust and non-robust models, show that $\sigma$\texttt{-zero} finds minimum $\ell_0$-norm adversarial examples without requiring any time-consuming hyperparameter tuning, and that it outperforms all competing sparse attacks in terms of success rate, perturbation size, and efficiency.
Authors: Alhassan Mumuni, Fuseini Mumuni
Abstract: Data augmentation is arguably the most important regularization technique commonly used to improve generalization performance of machine learning models. It primarily involves the application of appropriate data transformation operations to create new data samples with desired properties. Despite its effectiveness, the process is often challenging because of the time-consuming trial and error procedures for creating and testing different candidate augmentations and their hyperparameters manually. State-of-the-art approaches are increasingly relying on automated machine learning (AutoML) principles. This work presents a comprehensive survey of AutoML-based data augmentation techniques. We discuss various approaches for accomplishing data augmentation with AutoML, including data manipulation, data integration and data synthesis techniques. The focus of this work is on image data augmentation methods. Nonetheless, we cover other data modalities, especially in cases where the specific data augmentations techniques being discussed are more suitable for these other modalities. For instance, since automated data integration methods are more suitable for tabular data, we cover tabular data in the discussion of data integration methods. The work also presents extensive discussion of techniques for accomplishing each of the major subtasks of the image data augmentation process: search space design, hyperparameter optimization and model evaluation. Finally, we carried out an extensive comparison and analysis of the performance of automated data augmentation techniques and state-of-the-art methods based on classical augmentation approaches. The results show that AutoML methods for data augmentation currently outperform state-of-the-art techniques based on conventional approaches.
Authors: Xiaoshuai Chen, Wei Chen, Dongmyoung Lee, Yukun Ge, Nicolas Rojas, Petar Kormushev
Abstract: End-to-end robot learning, particularly for long-horizon tasks, often results in unpredictable outcomes and poor generalization. To address these challenges, we propose a novel Therblig-Based Backbone Framework (TBBF) as a fundamental structure to enhance interpretability, data efficiency, and generalization in robotic systems. TBBF utilizes expert demonstrations to enable therblig-level task decomposition, facilitate efficient action-object mapping, and generate adaptive trajectories for new scenarios. The approach consists of two stages: offline training and online testing. During the offline training stage, we developed the Meta-RGate SynerFusion (MGSF) network for accurate therblig segmentation across various tasks. In the online testing stage, after a one-shot demonstration of a new task is collected, our MGSF network extracts high-level knowledge, which is then encoded into the image using Action Registration (ActionREG). Additionally, Large Language Model (LLM)-Alignment Policy for Visual Correction (LAP-VC) is employed to ensure precise action registration, facilitating trajectory transfer in novel robot scenarios. Experimental results validate these methods, achieving 94.37% recall in therblig segmentation and success rates of 94.4% and 80% in real-world online robot testing for simple and complex scenarios, respectively. Supplementary material is available at: https://sites.google.com/view/therbligsbasedbackbone/home
URLs: https://sites.google.com/view/therbligsbasedbackbone/home
Authors: Yongxing Zhang, Donglin Yang, Renjie Liao
Abstract: Finite symmetric groups $S_n$ are essential in fields such as combinatorics, physics, and chemistry. However, learning a probability distribution over $S_n$ poses significant challenges due to its intractable size and discrete nature. In this paper, we introduce SymmetricDiffusers, a novel discrete diffusion model that simplifies the task of learning a complicated distribution over $S_n$ by decomposing it into learning simpler transitions of the reverse diffusion using deep neural networks. We identify the riffle shuffle as an effective forward transition and provide empirical guidelines for selecting the diffusion length based on the theory of random walks on finite groups. Additionally, we propose a generalized Plackett-Luce (PL) distribution for the reverse transition, which is provably more expressive than the PL distribution. We further introduce a theoretically grounded "denoising schedule" to improve sampling and learning efficiency. Extensive experiments show that our model achieves state-of-the-art or comparable performances on solving tasks including sorting 4-digit MNIST images, jigsaw puzzles, and traveling salesman problems. Our code is released at https://github.com/DSL-Lab/SymmetricDiffusers.
Authors: Ayano Hiranaka, Shang-Fu Chen, Chieh-Hsin Lai, Dongjun Kim, Naoki Murata, Takashi Shibuya, Wei-Hsiang Liao, Shao-Hua Sun, Yuki Mitsufuji
Abstract: Controllable generation through Stable Diffusion (SD) fine-tuning aims to improve fidelity, safety, and alignment with human guidance. Existing reinforcement learning from human feedback methods usually rely on predefined heuristic reward functions or pretrained reward models built on large-scale datasets, limiting their applicability to scenarios where collecting such data is costly or difficult. To effectively and efficiently utilize human feedback, we develop a framework, HERO, which leverages online human feedback collected on the fly during model learning. Specifically, HERO features two key mechanisms: (1) Feedback-Aligned Representation Learning, an online training method that captures human feedback and provides informative learning signals for fine-tuning, and (2) Feedback-Guided Image Generation, which involves generating images from SD's refined initialization samples, enabling faster convergence towards the evaluator's intent. We demonstrate that HERO is 4x more efficient in online feedback for body part anomaly correction compared to the best existing method. Additionally, experiments show that HERO can effectively handle tasks like reasoning, counting, personalization, and reducing NSFW content with only 0.5K online feedback.
Authors: Jun-En Ding, Chien-Chin Hsu, Chi-Hsiang Chu, Shuqiang Wang, Feng Liu
Abstract: The classification of medical images is a pivotal aspect of disease diagnosis, often enhanced by deep learning techniques. However, traditional approaches typically focus on unimodal medical image data, neglecting the integration of diverse non-image patient data. This paper proposes a novel Cross-Graph Modal Contrastive Learning (CGMCL) framework for multimodal structured data from different data domains to improve medical image classification. The model effectively integrates both image and non-image data by constructing cross-modality graphs and leveraging contrastive learning to align multimodal features in a shared latent space. An inter-modality feature scaling module further optimizes the representation learning process by reducing the gap between heterogeneous modalities. The proposed approach is evaluated on two datasets: a Parkinson's disease (PD) dataset and a public melanoma dataset. Results demonstrate that CGMCL outperforms conventional unimodal methods in accuracy, interpretability, and early disease prediction. Additionally, the method shows superior performance in multi-class melanoma classification. The CGMCL framework provides valuable insights into medical image classification while offering improved disease interpretability and predictive capabilities.
Authors: Till Aczel, Roger Wattenhofer
Abstract: In lossy image compression, models face the challenge of either hallucinating details or generating out-of-distribution samples due to the information bottleneck. This implies that at times, introducing hallucinations is necessary to generate in-distribution samples. The optimal level of hallucination varies depending on image content, as humans are sensitive to small changes that alter the semantic meaning. We propose a novel compression method that dynamically balances the degree of hallucination based on content. We collect data and train a model to predict user preferences on hallucinations. By using this prediction to adjust the perceptual weight in the reconstruction loss, we develop a Conditionally Hallucinating compression model (ConHa) that outperforms state-of-the-art image compression methods. Code and images are available at https://polybox.ethz.ch/index.php/s/owS1k5JYs4KD4TA.
Authors: Zhenyu Jiang, Yuqi Xie, Kevin Lin, Zhenjia Xu, Weikang Wan, Ajay Mandlekar, Linxi Fan, Yuke Zhu
Abstract: Imitation learning from human demonstrations is an effective means to teach robots manipulation skills. But data acquisition is a major bottleneck in applying this paradigm more broadly, due to the amount of cost and human effort involved. There has been significant interest in imitation learning for bimanual dexterous robots, like humanoids. Unfortunately, data collection is even more challenging here due to the challenges of simultaneously controlling multiple arms and multi-fingered hands. Automated data generation in simulation is a compelling, scalable alternative to fuel this need for data. To this end, we introduce DexMimicGen, a large-scale automated data generation system that synthesizes trajectories from a handful of human demonstrations for humanoid robots with dexterous hands. We present a collection of simulation environments in the setting of bimanual dexterous manipulation, spanning a range of manipulation behaviors and different requirements for coordination among the two arms. We generate 21K demos across these tasks from just 60 source human demos and study the effect of several data generation and policy learning decisions on agent performance. Finally, we present a real-to-sim-to-real pipeline and deploy it on a real-world humanoid can sorting task. Generated datasets, simulation environments and additional results are at https://dexmimicgen.github.io/
Authors: Yangyang Guo, Fangkai Jiao, Liqiang Nie, Mohan Kankanhalli
Abstract: The vulnerability of Vision Large Language Models (VLLMs) to jailbreak attacks appears as no surprise. However, recent defense mechanisms against these attacks have reached near-saturation performance on benchmark evaluations, often with minimal effort. This \emph{dual high performance} in both attack and defense raises a fundamental and perplexing paradox. To gain a deep understanding of this issue and thus further help strengthen the trustworthiness of VLLMs, this paper makes three key contributions: i) One tentative explanation for VLLMs being prone to jailbreak attacks--\textbf{inclusion of vision inputs}, as well as its in-depth analysis. ii) The recognition of a largely ignored problem in existing defense mechanisms--\textbf{over-prudence}. The problem causes these defense methods to exhibit unintended abstention, even in the presence of benign inputs, thereby undermining their reliability in faithfully defending against attacks. iii) A simple safety-aware method--\textbf{LLM-Pipeline}. Our method repurposes the more advanced guardrails of LLMs on the shelf, serving as an effective alternative detector prior to VLLM response. Last but not least, we find that the two representative evaluation methods for jailbreak often exhibit chance agreement. This limitation makes it potentially misleading when evaluating attack strategies or defense mechanisms. We believe the findings from this paper offer useful insights to rethink the foundational development of VLLM safety with respect to benchmark datasets, defense strategies, and evaluation methods.
Authors: Hu Wang, Congbo Ma, Ibrahim Almakky, Ian Reid, Gustavo Carneiro, Mohammad Yaqub
Abstract: Model-merging has emerged as a powerful approach in deep learning, capable of enhancing model performance without any training. However, the underlying mechanisms that explain its effectiveness remain largely unexplored. In this paper, we investigate this technique from three novel perspectives to empirically provide deeper insights into why and how weight-averaged model-merging works: (1) we examine the intrinsic patterns captured by the learning of the model weights, through the visualizations of their patterns on several datasets, showing that these weights often encode structured and interpretable patterns and that is the essential why model-merging can work; (2) we mathematically and empirically investigate model ensemble merging strategies based on averaging on weights versus averaging on features, providing detailed analyses across diverse architectures and datasets; and (3) we explore the impact on model-merging prediction stability in terms of changing the parameter magnitude, revealing insights into the way of weight averaging works as regularization by showing the robustness across different parameter scales. Our findings shed light on the "black box" of weight-averaged model-merging, offering valuable insights and practical recommendations that advance the model-merging process. The code is available at https://github.com/billhhh/Rethink-Merge.
Authors: Lei Li, Sen Jia, Wang Jianhao, Zhongyu Jiang, Feng Zhou, Ju Dai, Tianfang Zhang, Wu Zongkai, Jenq-Neng Hwang
Abstract: This paper presents LLaMo (Large Language and Human Motion Assistant), a multimodal framework for human motion instruction tuning. In contrast to conventional instruction-tuning approaches that convert non-linguistic inputs, such as video or motion sequences, into language tokens, LLaMo retains motion in its native form for instruction tuning. This method preserves motion-specific details that are often diminished in tokenization, thereby improving the model's ability to interpret complex human behaviors. By processing both video and motion data alongside textual inputs, LLaMo enables a flexible, human-centric analysis. Experimental evaluations across high-complexity domains, including human behaviors and professional activities, indicate that LLaMo effectively captures domain-specific knowledge, enhancing comprehension and prediction in motion-intensive scenarios. We hope LLaMo offers a foundation for future multimodal AI systems with broad applications, from sports analytics to behavioral prediction. Our code and models are available on the project website: https://github.com/ILGLJ/LLaMo.
Authors: Zhen Liu, Tim Z. Xiao, Weiyang Liu, Yoshua Bengio, Dinghuai Zhang
Abstract: While one commonly trains large diffusion models by collecting datasets on target downstream tasks, it is often desired to align and finetune pretrained diffusion models with some reward functions that are either designed by experts or learned from small-scale datasets. Existing post-training methods for reward finetuning of diffusion models typically suffer from lack of diversity in generated samples, lack of prior preservation, and/or slow convergence in finetuning. Inspired by recent successes in generative flow networks (GFlowNets), a class of probabilistic models that sample with the unnormalized density of a reward function, we propose a novel GFlowNet method dubbed Nabla-GFlowNet (abbreviated as \methodname), the first GFlowNet method that leverages the rich signal in reward gradients, together with an objective called \graddb plus its variant \resgraddb designed for prior-preserving diffusion finetuning. We show that our proposed method achieves fast yet diversity- and prior-preserving finetuning of Stable Diffusion, a large-scale text-conditioned image diffusion model, on different realistic reward functions.
Authors: Weizhi Xian, Mingliang Zhou, Leong Hou U, Lang Shujun, Bin Fang, Tao Xiang, Zhaowei Shang, Weijia Jia
Abstract: In this paper, we propose a Physical Imaging Guided perceptual framework for Underwater Image Quality Assessment (UIQA), termed PIGUIQA. First, we formulate UIQA as a comprehensive problem that considers the combined effects of direct transmission attenuation and backward scattering on image perception. By leveraging underwater radiative transfer theory, we systematically integrate physics-based imaging estimations to establish quantitative metrics for these distortions. Second, recognizing spatial variations in image content significance and human perceptual sensitivity to distortions, we design a module built upon a neighborhood attention mechanism for local perception of images. This module effectively captures subtle features in images, thereby enhancing the adaptive perception of distortions on the basis of local information. Third, by employing a global perceptual aggregator that further integrates holistic image scene with underwater distortion information, the proposed model accurately predicts image quality scores. Extensive experiments across multiple benchmarks demonstrate that PIGUIQA achieves state-of-the-art performance while maintaining robust cross-dataset generalizability. The implementation is publicly available at https://anonymous.4open.science/r/PIGUIQA-A465/
Authors: Young Seok Jeon, Hongfei Yang, Huazhu Fu, Mengling Feng
Abstract: 3D models surpass 2D models in CT/MRI segmentation by effectively capturing inter-slice relationships. However, the added depth dimension substantially increases memory consumption. While patch-based training alleviates memory constraints, it significantly slows down the inference speed due to the sliding window (SW) approach. We propose No-More-Sliding-Window (NMSW), a novel end-to-end trainable framework that enhances the efficiency of generic 3D segmentation backbone during an inference step by eliminating the need for SW. NMSW employs a differentiable Top-k module to selectively sample only the most relevant patches, thereby minimizing redundant computations. When patch-level predictions are insufficient, the framework intelligently leverages coarse global predictions to refine results. Evaluated across 3 tasks using 3 segmentation backbones, NMSW achieves competitive accuracy compared to SW inference while significantly reducing computational complexity by 91% (88.0 to 8.00 TMACs). Moreover, it delivers a 9.1x faster inference on the H100 GPU (99.0 to 8.3 sec) and a 11.1x faster inference on the Xeon Gold CPU (2110 to 189 sec). NMSW is model-agnostic, further boosting efficiency when integrated with any existing efficient segmentation backbones.
Authors: Lucas Lange, Ole Borchardt, Erhard Rahm
Abstract: With growing concerns about user data collection, individualized privacy has emerged as a promising solution to balance protection and utility by accounting for diverse user privacy preferences. Instead of enforcing a uniform level of anonymization for all users, this approach allows individuals to choose privacy settings that align with their comfort levels. Building on this idea, we propose an adapted method for enabling Individualized Differential Privacy (IDP) in Federated Learning (FL) by handling clients according to their personal privacy preferences. By extending the SAMPLE algorithm from centralized settings to FL, we calculate client-specific sampling rates based on their heterogeneous privacy budgets and integrate them into a modified IDP-FedAvg algorithm. We test this method under realistic privacy distributions and multiple datasets. The experimental results demonstrate that our approach achieves clear improvements over uniform DP baselines, reducing the trade-off between privacy and utility. Compared to the alternative SCALE method in related work, which assigns differing noise scales to clients, our method performs notably better. However, challenges remain for complex tasks with non-i.i.d. data, primarily stemming from the constraints of the decentralized setting.
Authors: Yansong Qu, Dian Chen, Xinyang Li, Xiaofan Li, Shengchuan Zhang, Liujuan Cao, Rongrong Ji
Abstract: Recent advancements in 3D scene editing have been propelled by the rapid development of generative models. Existing methods typically utilize generative models to perform text-guided editing on 3D representations, such as 3D Gaussian Splatting (3DGS). However, these methods are often limited to texture modifications and fail when addressing geometric changes, such as editing a character's head to turn around. Moreover, such methods lack accurate control over the spatial position of editing results, as language struggles to precisely describe the extent of edits. To overcome these limitations, we introduce DYG, an effective 3D drag-based editing method for 3D Gaussian Splatting. It enables users to conveniently specify the desired editing region and the desired dragging direction through the input of 3D masks and pairs of control points, thereby enabling precise control over the extent of editing. DYG integrates the strengths of the implicit triplane representation to establish the geometric scaffold of the editing results, effectively overcoming suboptimal editing outcomes caused by the sparsity of 3DGS in the desired editing regions. Additionally, we incorporate a drag-based Latent Diffusion Model into our method through the proposed Drag-SDS loss function, enabling flexible, multi-view consistent, and fine-grained editing. Extensive experiments demonstrate that DYG conducts effective drag-based editing guided by control point prompts, surpassing other baselines in terms of editing effect and quality, both qualitatively and quantitatively. Visit our project page at https://quyans.github.io/Drag-Your-Gaussian.
Authors: Xiaoya Lu, Dongrui Liu, Yi Yu, Luxin Xu, Jing Shao
Abstract: Despite the rapid development of safety alignment techniques for LLMs, defending against multi-turn jailbreaks is still a challenging task. In this paper, we conduct a comprehensive comparison, revealing that some existing defense methods can improve the robustness of LLMs against multi-turn jailbreaks but compromise usability, i.e., reducing general capabilities or causing the over-refusal problem. From the perspective of mechanism interpretability of LLMs, we discover that these methods fail to establish a boundary that exactly distinguishes safe and harmful feature representations. Therefore, boundary-safe representations close to harmful representations are inevitably disrupted, leading to a decline in usability. To address this issue, we propose X-Boundary to push harmful representations away from boundary-safe representations and obtain an exact distinction boundary. In this way, harmful representations can be precisely erased without disrupting safe ones. Experimental results show that X-Boundary achieves state-of-the-art defense performance against multi-turn jailbreaks, while reducing the over-refusal rate by about 20% and maintaining nearly complete general capability. Furthermore, we theoretically prove and empirically verify that X-Boundary can accelerate the convergence process during training. Please see our code at: https://github.com/AI45Lab/X-Boundary.
Authors: Eldad Haber, Shadab Ahamed, Md. Shahriar Rahim Siddiqui, Niloufar Zakariaei, Moshe Eliasof
Abstract: Generative models for image generation are now commonly used for a wide variety of applications, ranging from guided image generation for entertainment to solving inverse problems. Nonetheless, training a generator is a non-trivial feat that requires fine-tuning and can lead to so-called hallucinations, that is, the generation of images that are unrealistic. In this work, we explore image generation using flow matching. We explain and demonstrate why flow matching can generate hallucinations, and propose an iterative process to improve the generation process. Our iterative process can be integrated into virtually $\textit{any}$ generative modeling technique, thereby enhancing the performance and robustness of image synthesis systems.
Authors: Jun Zeng, Debesh Jha, Ertugrul Aktas, Elif Keles, Alpay Medetalibeyoglu, Matthew Antalek, Robert Lewandowski, Daniela Ladner, Amir A. Borhani, Gorkem Durak, Ulas Bagci
Abstract: We present RMA-Mamba, a novel architecture that advances the capabilities of vision state space models through a specialized reverse mamba attention module (RMA). The key innovation lies in RMA-Mamba's ability to capture long-range dependencies while maintaining precise local feature representation through its hierarchical processing pipeline. By integrating Vision Mamba (VMamba)'s efficient sequence modeling with RMA's targeted feature refinement, our architecture achieves superior feature learning across multiple scales. This dual-mechanism approach enables robust handling of complex morphological patterns while maintaining computational efficiency. We demonstrate RMA-Mamba's effectiveness in the challenging domain of pathological liver segmentation (from both CT and MRI), where traditional segmentation approaches often fail due to tissue variations. When evaluated on a newly introduced cirrhotic liver dataset (CirrMRI600+) of T2-weighted MRI scans, RMA-Mamba achieves the state-of-the-art performance with a Dice coefficient of 92.08%, mean IoU of 87.36%, and recall of 92.96%. The architecture's generalizability is further validated on the cancerous liver segmentation from CT scans (LiTS: Liver Tumor Segmentation dataset), yielding a Dice score of 92.9% and mIoU of 88.99%. Our code is available for public: https://github.com/JunZengz/RMAMamba.
Authors: Henrui Tian, Wenhui Lei, Linrui Dai, Hanyu Chen, Xiaofan Zhang
Abstract: Fully-supervised lesion recognition methods in medical imaging face challenges due to the reliance on large annotated datasets, which are expensive and difficult to collect. To address this, synthetic lesion generation has become a promising approach. However, existing models struggle with scalability, fine-grained control over lesion attributes, and the generation of complex structures. We propose LesionDiffusion, a text-controllable lesion synthesis framework for 3D CT imaging that generates both lesions and corresponding masks. By utilizing a structured lesion report template, our model provides greater control over lesion attributes and supports a wider variety of lesion types. We introduce a dataset of 1,505 annotated CT scans with paired lesion masks and structured reports, covering 14 lesion types across 8 organs. LesionDiffusion consists of two components: a lesion mask synthesis network (LMNet) and a lesion inpainting network (LINet), both guided by lesion attributes and image features. Extensive experiments demonstrate that LesionDiffusion significantly improves segmentation performance, with strong generalization to unseen lesion types and organs, outperforming current state-of-the-art models. Code will be available at https://github.com/HengruiTianSJTU/LesionDiffusion.
Authors: Shashank Gupta, Chaitanya Ahuja, Tsung-Yu Lin, Sreya Dutta Roy, Harrie Oosterhuis, Maarten de Rijke, Satya Narayan Shukla
Abstract: Reinforcement learning (RL)-based fine-tuning has emerged as a powerful approach for aligning diffusion models with black-box objectives. Proximal policy optimization (PPO) is the most popular choice of method for policy optimization. While effective in terms of performance, PPO is highly sensitive to hyper-parameters and involves substantial computational overhead. REINFORCE, on the other hand, mitigates some computational complexities such as high memory overhead and sensitive hyper-parameter tuning, but has suboptimal performance due to high-variance and sample inefficiency. While the variance of the REINFORCE can be reduced by sampling multiple actions per input prompt and using a baseline correction term, it still suffers from sample inefficiency. To address these challenges, we systematically analyze the efficiency-effectiveness trade-off between REINFORCE and PPO, and propose leave-one-out PPO (LOOP), a novel RL for diffusion fine-tuning method. LOOP combines variance reduction techniques from REINFORCE, such as sampling multiple actions per input prompt and a baseline correction term, with the robustness and sample efficiency of PPO via clipping and importance sampling. Our results demonstrate that LOOP effectively improves diffusion models on various black-box objectives, and achieves a better balance between computational efficiency and performance.