Authors: Yongjia Ma, Donglin Di, Xuan Liu, Xiaokai Chen, Lei Fan, Wei Chen, Tonghua Su
Abstract: Rectified flow models have achieved remarkable performance in image and video generation tasks. However, existing numerical solvers face a trade-off between fast sampling and high-accuracy solutions, limiting their effectiveness in downstream applications such as reconstruction and editing. To address this challenge, we propose leveraging the Adams-Bashforth-Moulton (ABM) predictor-corrector method to enhance the accuracy of ODE solving in rectified flow models. Specifically, we introduce ABM-Solver, which integrates a multi step predictor corrector approach to reduce local truncation errors and employs Adaptive Step Size Adjustment to improve sampling speed. Furthermore, to effectively preserve non edited regions while facilitating semantic modifications, we introduce a Mask Guided Feature Injection module. We estimate self-similarity to generate a spatial mask that differentiates preserved regions from those available for editing. Extensive experiments on multiple high-resolution image datasets validate that ABM-Solver significantly improves inversion precision and editing quality, outperforming existing solvers without requiring additional training or optimization.
Authors: Jinchang Zhang, Guoyu Lu
Abstract: Depth estimation is a core problem in robotic perception and vision tasks, but 3D reconstruction from a single image presents inherent uncertainties. Current depth estimation models primarily rely on inter-image relationships for supervised training, often overlooking the intrinsic information provided by the camera itself. We propose a method that embodies the camera model and its physical characteristics into a deep learning model, computing embodied scene depth through real-time interactions with road environments. The model can calculate embodied scene depth in real-time based on immediate environmental changes using only the intrinsic properties of the camera, without any additional equipment. By combining embodied scene depth with RGB image features, the model gains a comprehensive perspective on both geometric and visual details. Additionally, we incorporate text descriptions containing environmental content and depth information as priors for scene understanding, enriching the model's perception of objects. This integration of image and language - two inherently ambiguous modalities - leverages their complementary strengths for monocular depth estimation. The real-time nature of the embodied language and depth prior model ensures that the model can continuously adjust its perception and behavior in dynamic environments. Experimental results show that the embodied depth estimation method enhances model performance across different scenes.
Authors: Bastian P\"atzold, Jan Nogga, Sven Behnke
Abstract: This paper introduces a novel approach that leverages the capabilities of vision-language models (VLMs) by integrating them with established approaches for open-vocabulary detection (OVD), instance segmentation, and tracking. We utilize VLM-generated structured descriptions to identify visible object instances, collect application-relevant attributes, and inform an open-vocabulary detector to extract corresponding bounding boxes that are passed to a video segmentation model providing precise segmentation masks and tracking capabilities. Once initialized, this model can then directly extract segmentation masks, allowing processing of image streams in real time with minimal computational overhead. Tracks can be updated online as needed by generating new structured descriptions and corresponding open-vocabulary detections. This combines the descriptive power of VLMs with the grounding capability of OVD and the pixel-level understanding and speed of video segmentation. Our evaluation across datasets and robotics platforms demonstrates the broad applicability of this approach, showcasing its ability to extract task-specific attributes from non-standard objects in dynamic environments.
Authors: Shiyi Jiang, Farshad Firouzi, Krishnendu Chakrabarty
Abstract: The increasing need for sharing healthcare data and collaborating on clinical research has raised privacy concerns. Health information leakage due to malicious attacks can lead to serious problems such as misdiagnoses and patient identification issues. Privacy-preserving machine learning (PPML) and privacy-enhancing technologies, particularly federated learning (FL), have emerged in recent years as innovative solutions to balance privacy protection with data utility; however, they also suffer from inherent privacy vulnerabilities. Gradient inversion attacks constitute major threats to data sharing in federated learning. Researchers have proposed many defenses against gradient inversion attacks. However, current defense methods for healthcare data lack generalizability, i.e., existing solutions may not be applicable to data from a broader range of populations. In addition, most existing defense methods are tested using non-healthcare data, which raises concerns about their applicability to real-world healthcare systems. In this study, we present a defense against gradient inversion attacks in federated learning. We achieve this using latent data perturbation and minimax optimization, utilizing both general and medical image datasets. Our method is compared to two baselines, and the results show that our approach can outperform the baselines with a reduction of 12.5% in the attacker's accuracy in classifying reconstructed images. The proposed method also yields an increase of over 12.4% in Mean Squared Error (MSE) between the original and reconstructed images at the same level of model utility of around 90% client classification accuracy. The results suggest the potential of a generalizable defense for healthcare data.
Authors: Saddam Hussain Khan (Artificial Intelligence Lab, Department of Computer Systems Engineering, University of Engineering,Applied Sciences), Rashid Iqbal (Artificial Intelligence Lab, Department of Computer Systems Engineering, University of Engineering,Applied Sciences)
Abstract: Deep Convolutional Neural Networks (CNNs) have significantly advanced deep learning, driving breakthroughs in computer vision, natural language processing, medical diagnosis, object detection, and speech recognition. Architectural innovations including 1D, 2D, and 3D convolutional models, dilated and grouped convolutions, depthwise separable convolutions, and attention mechanisms address domain-specific challenges and enhance feature representation and computational efficiency. Structural refinements such as spatial-channel exploitation, multi-path design, and feature-map enhancement contribute to robust hierarchical feature extraction and improved generalization, particularly through transfer learning. Efficient preprocessing strategies, including Fourier transforms, structured transforms, low-precision computation, and weight compression, optimize inference speed and facilitate deployment in resource-constrained environments. This survey presents a unified taxonomy that classifies CNN architectures based on spatial exploitation, multi-path structures, depth, width, dimensionality expansion, channel boosting, and attention mechanisms. It systematically reviews CNN applications in face recognition, pose estimation, action recognition, text classification, statistical language modeling, disease diagnosis, radiological analysis, cryptocurrency sentiment prediction, 1D data processing, video analysis, and speech recognition. In addition to consolidating architectural advancements, the review highlights emerging learning paradigms such as few-shot, zero-shot, weakly supervised, federated learning frameworks and future research directions include hybrid CNN-transformer models, vision-language integration, generative learning, etc. This review provides a comprehensive perspective on CNN's evolution from 2015 to 2025, outlining key innovations, challenges, and opportunities.
Authors: Felix Chen, Hangjie Yuan, Yunqiu Xu, Tao Feng, Jun Cen, Pengwei Liu, Zeying Huang, Yi Yang
Abstract: Despite impressive performance across diverse tasks, Multimodal Large Language Models (MLLMs) have yet to fully demonstrate their potential in visual mathematical problem-solving, particularly in accurately perceiving and interpreting diagrams. Inspired by typical processes of humans, we hypothesize that the perception capabilities to extract meaningful information from diagrams is crucial, as it directly impacts subsequent inference processes. To validate this hypothesis, we developed FlowVerse, a comprehensive benchmark that categorizes all information used during problem-solving into four components, which are then combined into six problem versions for evaluation. Our preliminary results on FlowVerse reveal that existing MLLMs exhibit substantial limitations when extracting essential information and reasoned property from diagrams and performing complex reasoning based on these visual inputs. In response, we introduce MathFlow, a modular problem-solving pipeline that decouples perception and inference into distinct stages, thereby optimizing each independently. Given the perceptual limitations observed in current MLLMs, we trained MathFlow-P-7B as a dedicated perception model. Experimental results indicate that MathFlow-P-7B yields substantial performance gains when integrated with various closed-source and open-source inference models. This demonstrates the effectiveness of the MathFlow pipeline and its compatibility to diverse inference frameworks. The FlowVerse benchmark and code are available at https://github.com/MathFlow-zju/MathFlow.
Authors: Jie Zhang, Zheng Yuan, Zhongqi Wang, Bei Yan, Sibo Wang, Xiangkui Cao, Zonghui Guo, Shiguang Shan, Xilin Chen
Abstract: The rapid evolution of Large Vision-Language Models (LVLMs) has highlighted the necessity for comprehensive evaluation frameworks that assess these models across diverse dimensions. While existing benchmarks focus on specific aspects such as perceptual abilities, cognitive capabilities, and safety against adversarial attacks, they often lack the breadth and depth required to provide a holistic understanding of LVLMs' strengths and limitations. To address this gap, we introduce REVAL, a comprehensive benchmark designed to evaluate the \textbf{RE}liability and \textbf{VAL}ue of LVLMs. REVAL encompasses over 144K image-text Visual Question Answering (VQA) samples, structured into two primary sections: Reliability, which assesses truthfulness (\eg, perceptual accuracy and hallucination tendencies) and robustness (\eg, resilience to adversarial attacks, typographic attacks, and image corruption), and Values, which evaluates ethical concerns (\eg, bias and moral understanding), safety issues (\eg, toxicity and jailbreak vulnerabilities), and privacy problems (\eg, privacy awareness and privacy leakage). We evaluate 26 models, including mainstream open-source LVLMs and prominent closed-source models like GPT-4o and Gemini-1.5-Pro. Our findings reveal that while current LVLMs excel in perceptual tasks and toxicity avoidance, they exhibit significant vulnerabilities in adversarial scenarios, privacy preservation, and ethical reasoning. These insights underscore critical areas for future improvements, guiding the development of more secure, reliable, and ethically aligned LVLMs. REVAL provides a robust framework for researchers to systematically assess and compare LVLMs, fostering advancements in the field.
Authors: Jonas Krumme, Christoph Zetzsche
Abstract: When interacting with the world robots face a number of difficult questions, having to make decisions when given under-specified tasks where they need to make choices, often without clearly defined right and wrong answers. Humans, on the other hand, can often rely on their knowledge and experience to fill in the gaps. For example, the simple task of organizing newly bought produce into the fridge involves deciding where to put each thing individually, how to arrange them together meaningfully, e.g. putting related things together, all while there is no clear right and wrong way to accomplish this task. We could encode all this information on how to do such things explicitly into the robots' knowledge base, but this can quickly become overwhelming, considering the number of potential tasks and circumstances the robot could encounter. However, images of the real world often implicitly encode answers to such questions and can show which configurations of objects are meaningful or are usually used by humans. An image of a full fridge can give a lot of information about how things are usually arranged in relation to each other and the full fridge at large. Modern generative systems are capable of generating plausible images of the real world and can be conditioned on the environment in which the robot operates. Here we investigate the idea of using the implicit knowledge about the world of modern generative AI systems given by their ability to generate convincing images of the real world to solve under-specified tasks.
Authors: Luigi Piccinelli, Christos Sakaridis, Mattia Segu, Yung-Hsu Yang, Siyuan Li, Wim Abbeloos, Luc Van Gool
Abstract: Monocular 3D estimation is crucial for visual perception. However, current methods fall short by relying on oversimplified assumptions, such as pinhole camera models or rectified images. These limitations severely restrict their general applicability, causing poor performance in real-world scenarios with fisheye or panoramic images and resulting in substantial context loss. To address this, we present UniK3D, the first generalizable method for monocular 3D estimation able to model any camera. Our method introduces a spherical 3D representation which allows for better disentanglement of camera and scene geometry and enables accurate metric 3D reconstruction for unconstrained camera models. Our camera component features a novel, model-independent representation of the pencil of rays, achieved through a learned superposition of spherical harmonics. We also introduce an angular loss, which, together with the camera module design, prevents the contraction of the 3D outputs for wide-view cameras. A comprehensive zero-shot evaluation on 13 diverse datasets demonstrates the state-of-the-art performance of UniK3D across 3D, depth, and camera metrics, with substantial gains in challenging large-field-of-view and panoramic settings, while maintaining top accuracy in conventional pinhole small-field-of-view domains. Code and models are available at github.com/lpiccinelli-eth/unik3d .
Authors: Katja Schwarz, Denys Rozumnyi, Samuel Rota Bul\`o, Lorenzo Porzi, Peter Kontschieder
Abstract: We introduce a recipe for generating immersive 3D worlds from a single image by framing the task as an in-context learning problem for 2D inpainting models. This approach requires minimal training and uses existing generative models. Our process involves two steps: generating coherent panoramas using a pre-trained diffusion model and lifting these into 3D with a metric depth estimator. We then fill unobserved regions by conditioning the inpainting model on rendered point clouds, requiring minimal fine-tuning. Tested on both synthetic and real images, our method produces high-quality 3D environments suitable for VR display. By explicitly modeling the 3D structure of the generated environment from the start, our approach consistently outperforms state-of-the-art, video synthesis-based methods along multiple quantitative image quality metrics. Project Page: https://katjaschwarz.github.io/worlds/
Authors: Xiaoran Zhang, Byung-Woo Hong, Hyoungseob Park, Daniel H. Pak, Anne-Marie Rickmann, Lawrence H. Staib, James S. Duncan, Alex Wong
Abstract: We propose a model-agnostic, progressive test-time energy adaptation approach for medical image segmentation. Maintaining model performance across diverse medical datasets is challenging, as distribution shifts arise from inconsistent imaging protocols and patient variations. Unlike domain adaptation methods that require multiple passes through target data - impractical in clinical settings - our approach adapts pretrained models progressively as they process test data. Our method leverages a shape energy model trained on source data, which assigns an energy score at the patch level to segmentation maps: low energy represents in-distribution (accurate) shapes, while high energy signals out-of-distribution (erroneous) predictions. By minimizing this energy score at test time, we refine the segmentation model to align with the target distribution. To validate the effectiveness and adaptability, we evaluated our framework on eight public MRI (bSSFP, T1- and T2-weighted) and X-ray datasets spanning cardiac, spinal cord, and lung segmentation. We consistently outperform baselines both quantitatively and qualitatively.
Authors: Moshiur Rahman Tonmoy, Md. Mithun Hossain, Nilanjan Dey, M. F. Mridha
Abstract: Plant diseases significantly threaten global food security by reducing crop yields and undermining agricultural sustainability. AI-driven automated classification has emerged as a promising solution, with deep learning models demonstrating impressive performance in plant disease identification. However, deploying these models on mobile and edge devices remains challenging due to high computational demands and resource constraints, highlighting the need for lightweight, accurate solutions for accessible smart agriculture systems. To address this, we propose MobilePlantViT, a novel hybrid Vision Transformer (ViT) architecture designed for generalized plant disease classification, which optimizes resource efficiency while maintaining high performance. Extensive experiments across diverse plant disease datasets of varying scales show our model's effectiveness and strong generalizability, achieving test accuracies ranging from 80% to over 99%. Notably, with only 0.69 million parameters, our architecture outperforms the smallest versions of MobileViTv1 and MobileViTv2, despite their higher parameter counts. These results underscore the potential of our approach for real-world, AI-powered automated plant disease classification in sustainable and resource-efficient smart agriculture systems. All codes will be available in the GitHub repository: https://github.com/moshiurtonmoy/MobilePlantViT
Authors: Hanxiao Wang, Biao Zhang, Weize Quan, Dong-Ming Yan, Peter Wonka
Abstract: This paper propose iFlame, a novel transformer-based network architecture for mesh generation. While attention-based models have demonstrated remarkable performance in mesh generation, their quadratic computational complexity limits scalability, particularly for high-resolution 3D data. Conversely, linear attention mechanisms offer lower computational costs but often struggle to capture long-range dependencies, resulting in suboptimal outcomes. To address this trade-off, we propose an interleaving autoregressive mesh generation framework that combines the efficiency of linear attention with the expressive power of full attention mechanisms. To further enhance efficiency and leverage the inherent structure of mesh representations, we integrate this interleaving approach into an hourglass architecture, which significantly boosts efficiency. Our approach reduces training time while achieving performance comparable to pure attention-based models. To improve inference efficiency, we implemented a caching algorithm that almost doubles the speed and reduces the KV cache size by seven-eighths compared to the original Transformer. We evaluate our framework on ShapeNet and Objaverse, demonstrating its ability to generate high-quality 3D meshes efficiently. Our results indicate that the proposed interleaving framework effectively balances computational efficiency and generative performance, making it a practical solution for mesh generation. The training takes only 2 days with 4 GPUs on 39k data with a maximum of 4k faces on Objaverse.
Authors: Eduard Allakhverdov, Elizaveta Goncharova, Andrey Kuznetsov
Abstract: Vision encoders typically generate a large number of visual tokens, providing information-rich representations but significantly increasing computational demands. This raises the question of whether all generated tokens are equally valuable or if some of them can be discarded to reduce computational costs without compromising quality. In this paper, we introduce a new method for determining feature utility based on the idea that less valuable features can be reconstructed from more valuable ones. We implement this concept by integrating an autoencoder with a Gumbel-Softmax selection mechanism, that allows identifying and retaining only the most informative visual tokens. To validate our approach, we compared the performance of the LLaVA-NeXT model, using features selected by our method with randomly selected features. We found that on OCR-based tasks, more than 50% of the visual context can be removed with minimal performance loss, whereas randomly discarding the same proportion of features significantly affects the model capabilities. Furthermore, in general-domain tasks, even randomly retaining only 30% of tokens achieves performance comparable to using the full set of visual tokens. Our results highlight a promising direction towards adaptive and efficient multimodal pruning that facilitates scalable and low-overhead inference without compromising performance.
Authors: Martin Kosteln\'ik, Karel Bene\v{s}, Michal Hradi\v{s}
Abstract: Logical page segmentation is an important step in document analysis, enabling better semantic representations, information retrieval, and text understanding. Previous approaches define logical segmentation either through text or geometric objects, relying on OCR or precise geometry. To avoid the need for OCR, we define the task purely as segmentation in the image domain. Furthermore, to ensure the evaluation remains unaffected by geometrical variations that do not impact text segmentation, we propose to use only foreground text pixels in the evaluation metric and disregard all background pixels. To support research in logical document segmentation, we introduce TextBite, a dataset of historical Czech documents spanning the 18th to 20th centuries, featuring diverse layouts from newspapers, dictionaries, and handwritten records. The dataset comprises 8,449 page images with 78,863 annotated segments of logically and thematically coherent text. We propose a set of baseline methods combining text region detection and relation prediction. The dataset, baselines and evaluation framework can be accessed at https://github.com/DCGM/textbite-dataset.
Authors: Zeping Liu, Fan Zhang, Junfeng Jiao, Ni Lao, Gengchen Mai
Abstract: Advancements in vision and language foundation models have inspired the development of geo-foundation models (GeoFMs), enhancing performance across diverse geospatial tasks. However, many existing GeoFMs primarily focus on overhead remote sensing (RS) data while neglecting other data modalities such as ground-level imagery. A key challenge in multimodal GeoFM development is to explicitly model geospatial relationships across modalities, which enables generalizability across tasks, spatial scales, and temporal contexts. To address these limitations, we propose GAIR, a novel multimodal GeoFM architecture integrating overhead RS data, street view (SV) imagery, and their geolocation metadata. We utilize three factorized neural encoders to project an SV image, its geolocation, and an RS image into the embedding space. The SV image needs to be located within the RS image's spatial footprint but does not need to be at its geographic center. In order to geographically align the SV image and RS image, we propose a novel implicit neural representations (INR) module that learns a continuous RS image representation and looks up the RS embedding at the SV image's geolocation. Next, these geographically aligned SV embedding, RS embedding, and location embedding are trained with contrastive learning objectives from unlabeled data. We evaluate GAIR across 10 geospatial tasks spanning RS image-based, SV image-based, and location embedding-based benchmarks. Experimental results demonstrate that GAIR outperforms state-of-the-art GeoFMs and other strong baselines, highlighting its effectiveness in learning generalizable and transferable geospatial representations.
Authors: Jinlong Li, Cristiano Saltori, Fabio Poiesi, Nicu Sebe
Abstract: The lack of a large-scale 3D-text corpus has led recent works to distill open-vocabulary knowledge from vision-language models (VLMs). owever, these methods typically rely on a single VLM to align the feature spaces of 3D models within a common language space, which limits the potential of 3D models to leverage the diverse spatial and semantic capabilities encapsulated in various foundation models. In this paper, we propose Cross-modal and Uncertainty-aware Agglomeration for Open-vocabulary 3D Scene Understanding dubbed CUA-O3D, the first model to integrate multiple foundation models-such as CLIP, DINOv2, and Stable Diffusion-into 3D scene understanding. We further introduce a deterministic uncertainty estimation to adaptively distill and harmonize the heterogeneous 2D feature embeddings from these models. Our method addresses two key challenges: (1) incorporating semantic priors from VLMs alongside the geometric knowledge of spatially-aware vision foundation models, and (2) using a novel deterministic uncertainty estimation to capture model-specific uncertainties across diverse semantic and geometric sensitivities, helping to reconcile heterogeneous representations during training. Extensive experiments on ScanNetV2 and Matterport3D demonstrate that our method not only advances open-vocabulary segmentation but also achieves robust cross-domain alignment and competitive spatial perception capabilities. The code will be available at \href{https://github.com/TyroneLi/CUA_O3D}{CUA_O3D}.
Authors: Xuan Shen, Weize Ma, Jing Liu, Changdi Yang, Rui Ding, Quanyi Wang, Henghui Ding, Wei Niu, Yanzhi Wang, Pu Zhao, Jun Lin, Jiuxiang Gu
Abstract: Monocular Depth Estimation (MDE) has emerged as a pivotal task in computer vision, supporting numerous real-world applications. However, deploying accurate depth estimation models on resource-limited edge devices, especially Application-Specific Integrated Circuits (ASICs), is challenging due to the high computational and memory demands. Recent advancements in foundational depth estimation deliver impressive results but further amplify the difficulty of deployment on ASICs. To address this, we propose QuartDepth which adopts post-training quantization to quantize MDE models with hardware accelerations for ASICs. Our approach involves quantizing both weights and activations to 4-bit precision, reducing the model size and computation cost. To mitigate the performance degradation, we introduce activation polishing and compensation algorithm applied before and after activation quantization, as well as a weight reconstruction method for minimizing errors in weight quantization. Furthermore, we design a flexible and programmable hardware accelerator by supporting kernel fusion and customized instruction programmability, enhancing throughput and efficiency. Experimental results demonstrate that our framework achieves competitive accuracy while enabling fast inference and higher energy efficiency on ASICs, bridging the gap between high-performance depth estimation and practical edge-device applicability. Code: https://github.com/shawnricecake/quart-depth
Authors: Yanyan Li, Youxu Fang, Zunjie Zhu, Kunyi Li, Yong Ding, Federico Tombari
Abstract: Simultaneously localizing camera poses and constructing Gaussian radiance fields in dynamic scenes establish a crucial bridge between 2D images and the 4D real world. Instead of removing dynamic objects as distractors and reconstructing only static environments, this paper proposes an efficient architecture that incrementally tracks camera poses and establishes the 4D Gaussian radiance fields in unknown scenarios by using a sequence of RGB-D images. First, by generating motion masks, we obtain static and dynamic priors for each pixel. To eliminate the influence of static scenes and improve the efficiency on learning the motion of dynamic objects, we classify the Gaussian primitives into static and dynamic Gaussian sets, while the sparse control points along with an MLP is utilized to model the transformation fields of the dynamic Gaussians. To more accurately learn the motion of dynamic Gaussians, a novel 2D optical flow map reconstruction algorithm is designed to render optical flows of dynamic objects between neighbor images, which are further used to supervise the 4D Gaussian radiance fields along with traditional photometric and geometric constraints. In experiments, qualitative and quantitative evaluation results show that the proposed method achieves robust tracking and high-quality view synthesis performance in real-world environments.
Authors: Philipp Becker, Abhinav Mehrotra, Ruchika Chavhan, Malcolm Chadwick, Luca Morreale, Mehdi Noroozi, Alberto Gil Ramos, Sourav Bhattacharya
Abstract: Diffusion Transformers (DiTs) have emerged as a leading architecture for text-to-image synthesis, producing high-quality and photorealistic images. However, the quadratic scaling properties of the attention in DiTs hinder image generation with higher resolution or on devices with limited resources. This work introduces an efficient diffusion transformer (EDiT) to alleviate these efficiency bottlenecks in conventional DiTs and Multimodal DiTs (MM-DiTs). First, we present a novel linear compressed attention method that uses a multi-layer convolutional network to modulate queries with local information while keys and values are spatially aggregated. Second, we formulate a hybrid attention scheme for multi-modal inputs that combines linear attention for image-to-image interactions and standard scaled dot-product attention for interactions involving prompts. Merging these two approaches leads to an expressive, linear-time Multimodal Efficient Diffusion Transformer (MM-EDiT). We demonstrate the effectiveness of the EDiT and MM-EDiT architectures by integrating them into PixArt-Sigma(conventional DiT) and Stable Diffusion 3.5-Medium (MM-DiT), achieving up to 2.2x speedup with comparable image quality after distillation.
Authors: Esther Y. H. Lin, Yimin Ding, Jogendra Kundu, Yatong An, Mohamed T. El-Haddad, Alexander Fix
Abstract: Eye tracking (ET) is a key enabler for Augmented and Virtual Reality (AR/VR). Prototyping new ET hardware requires assessing the impact of hardware choices on eye tracking performance. This task is compounded by the high cost of obtaining data from sufficiently many variations of real hardware, especially for machine learning, which requires large training datasets. We propose a method for end-to-end evaluation of how hardware changes impact machine learning-based ET performance using only synthetic data. We utilize a dataset of real 3D eyes, reconstructed from light dome data using neural radiance fields (NeRF), to synthesize captured eyes from novel viewpoints and camera parameters. Using this framework, we demonstrate that we can predict the relative performance across various hardware configurations, accounting for variations in sensor noise, illumination brightness, and optical blur. We also compare our simulator with the publicly available eye tracking dataset from the Project Aria glasses, demonstrating a strong correlation with real-world performance. Finally, we present a first-of-its-kind analysis in which we vary ET camera positions, evaluating ET performance ranging from on-axis direct views of the eye to peripheral views on the frame. Such an analysis would have previously required manufacturing physical devices to capture evaluation data. In short, our method enables faster prototyping of ET hardware.
Authors: George Cazenavette, Joel Julin, Simon Lucey
Abstract: Until quite recently, the backbone of nearly every state-of-the-art computer vision model has been the 2D convolution. At its core, a 2D convolution simultaneously mixes information across both the spatial and channel dimensions of a representation. Many recent computer vision architectures consist of sequences of isotropic blocks that disentangle the spatial and channel-mixing components. This separation of the operations allows us to more closely juxtapose the effects of spatial and channel mixing in deep learning. In this paper, we take an initial step towards garnering a deeper understanding of the roles of these mixing operations. Through our experiments and analysis, we discover that on both classical (ResNet) and cutting-edge (ConvMixer) models, we can reach nearly the same level of classification performance by and leaving the spatial mixers at their random initializations. Furthermore, we show that models with random, fixed spatial mixing are naturally more robust to adversarial perturbations. Lastly, we show that this phenomenon extends past the classification regime, as such models can also decode pixel-shuffled images.
Authors: Meng Zhou, Jiadong Xie, Mingsheng Xu
Abstract: Mainstream visual object tracking frameworks predominantly rely on template matching paradigms. Their performance heavily depends on the quality of template features, which becomes increasingly challenging to maintain in complex scenarios involving target deformation, occlusion, and background clutter. While existing spatiotemporal memory-based trackers emphasize memory capacity expansion, they lack effective mechanisms for dynamic feature selection and adaptive fusion. To address this gap, we propose a Dynamic Attention Mechanism in Spatiotemporal Memory Network (DASTM) with two key innovations: 1) A differentiable dynamic attention mechanism that adaptively adjusts channel-spatial attention weights by analyzing spatiotemporal correlations between the templates and memory features; 2) A lightweight gating network that autonomously allocates computational resources based on target motion states, prioritizing high-discriminability features in challenging scenarios. Extensive evaluations on OTB-2015, VOT 2018, LaSOT, and GOT-10K benchmarks demonstrate our DASTM's superiority, achieving state-of-the-art performance in success rate, robustness, and real-time efficiency, thereby offering a novel solution for real-time tracking in complex environments.
Authors: Sreetama Sarkar, Sumit Bam Shrestha, Yue Che, Leobardo Campos-Macias, Gourav Datta, Peter A. Beerel
Abstract: The rapidly growing demand for on-chip edge intelligence on resource-constrained devices has motivated approaches to reduce energy and latency of deep learning models. Spiking neural networks (SNNs) have gained particular interest due to their promise to reduce energy consumption using event-based processing. We assert that while sigma-delta encoding in SNNs can take advantage of the temporal redundancy across video frames, they still involve a significant amount of redundant computations due to processing insignificant events. In this paper, we propose a region masking strategy that identifies regions of interest at the input of the SNN, thereby eliminating computation and data movement for events arising from unimportant regions. Our approach demonstrates that masking regions at the input not only significantly reduces the overall spiking activity of the network, but also provides significant improvement in throughput and latency. We apply region masking during video object detection on Loihi 2, demonstrating that masking approximately 60% of input regions can reduce energy-delay product by 1.65x over a baseline sigma-delta network, with a degradation in mAP@0.5 by 1.09%.
Authors: Valentin Bieri, Marco Zamboni, Nicolas S. Blumer, Qingxuan Chen, Francis Engelmann
Abstract: Vision-language models (VLMs) show great promise for 3D scene understanding but are mainly applied to indoor spaces or autonomous driving, focusing on low-level tasks like segmentation. This work expands their use to urban-scale environments by leveraging 3D reconstructions from multi-view aerial imagery. We propose OpenCity3D, an approach that addresses high-level tasks, such as population density estimation, building age classification, property price prediction, crime rate assessment, and noise pollution evaluation. Our findings highlight OpenCity3D's impressive zero-shot and few-shot capabilities, showcasing adaptability to new contexts. This research establishes a new paradigm for language-driven urban analytics, enabling applications in planning, policy, and environmental monitoring. See our project page: opencity3d.github.io
Authors: Uihyun Cho, Namhun Kim
Abstract: Recent advances in deep-learning based denoising methods have improved Low-Dose CT image quality. However, due to distinct HU distributions and diverse anatomical characteristics, a single model often struggles to generalize across multiple anatomies. To address this limitation, we introduce \textbf{Agent-Integrated Denoising Experts (A-IDE)} framework, which integrates three anatomical region-specialized RED-CNN models under the management of decision-making LLM agent. The agent analyzes semantic cues from BiomedCLIP to dynamically route incoming LDCT scans to the most appropriate expert model. We highlight three major advantages of our approach. A-IDE excels in heterogeneous, data-scarce environments. The framework automatically prevents overfitting by distributing tasks among multiple experts. Finally, our LLM-driven agentic pipeline eliminates the need for manual interventions. Experimental evaluations on the Mayo-2016 dataset confirm that A-IDE achieves superior performance in RMSE, PSNR, and SSIM compared to a single unified denoiser.
Authors: Enguang Wang, Zhimao Peng, Zhengyuan Xie, Haori Lu, Fei Yang, Xialei Liu
Abstract: Generalized Category Discovery (GCD) aims to classify unlabeled data containing both seen and novel categories. Although existing methods perform well on generic datasets, they struggle in fine-grained scenarios. We attribute this difficulty to their reliance on contrastive learning over global image features to automatically capture discriminative cues, which fails to capture the subtle local differences essential for distinguishing fine-grained categories. Therefore, in this paper, we propose incorporating part knowledge to address fine-grained GCD, which introduces two key challenges: the absence of annotations for novel classes complicates the extraction of the part features, and global contrastive learning prioritizes holistic feature invariance, inadvertently suppressing discriminative local part patterns. To address these challenges, we propose PartGCD, including 1) Adaptive Part Decomposition, which automatically extracts class-specific semantic parts via Gaussian Mixture Models, and 2) Part Discrepancy Regularization, enforcing explicit separation between part features to amplify fine-grained local part distinctions. Experiments demonstrate state-of-the-art performance across multiple fine-grained benchmarks while maintaining competitiveness on generic datasets, validating the effectiveness and robustness of our approach.
Authors: Haori Lu, Xusheng Cao, Linlan Huang, Enguang Wang, Fei Yang, Xialei Liu
Abstract: Continual learning aims to accumulate knowledge over a data stream while mitigating catastrophic forgetting. In Non-exemplar Class Incremental Learning (NECIL), forgetting arises during incremental optimization because old classes are inaccessible, hindering the retention of prior knowledge. To solve this, previous methods struggle in achieving the stability-plasticity balance in the training stages. However, we note that the testing stage is rarely considered among them, but is promising to be a solution to forgetting. Therefore, we propose RoSE, which is a simple yet effective method that \textbf{R}est\textbf{o}res forgotten knowledge through test-time \textbf{S}emantic \textbf{E}volution. Specifically designed for minimizing forgetting, RoSE is a test-time semantic drift compensation framework that enables more accurate drift estimation in a self-supervised manner. Moreover, to avoid incomplete optimization during online testing, we derive an analytical solution as an alternative to gradient descent. We evaluate RoSE on CIFAR-100, TinyImageNet, and ImageNet100 datasets, under both cold-start and warm-start settings. Our method consistently outperforms most state-of-the-art (SOTA) methods across various scenarios, validating the potential and feasibility of test-time evolution in NECIL.
Authors: Yihan Hu, Jianing Peng, Yiheng Lin, Ting Liu, Xiaochao Qu, Luoqi Liu, Yao Zhao, Yunchao Wei
Abstract: This paper presents a novel approach to improving text-guided image editing using diffusion-based models. Text-guided image editing task poses key challenge of precisly locate and edit the target semantic, and previous methods fall shorts in this aspect. Our method introduces a Precise Semantic Localization strategy that leverages visual and textual self-attention to enhance the cross-attention map, which can serve as a regional cues to improve editing performance. Then we propose a Dual-Level Control mechanism for incorporating regional cues at both feature and latent levels, offering fine-grained control for more precise edits. To fully compare our methods with other DiT-based approaches, we construct the RW-800 benchmark, featuring high resolution images, long descriptive texts, real-world images, and a new text editing task. Experimental results on the popular PIE-Bench and RW-800 benchmarks demonstrate the superior performance of our approach in preserving background and providing accurate edits.
Authors: Maoji Zheng, Ziyu Xu, Qiming Xia, Hai Wu, Chenglu Wen, Cheng Wang
Abstract: LiDAR-based 3D object detection and semantic segmentation are critical tasks in 3D scene understanding. Traditional detection and segmentation methods supervise their models through bounding box labels and semantic mask labels. However, these two independent labels inherently contain significant redundancy. This paper aims to eliminate the redundancy by supervising 3D object detection using only semantic labels. However, the challenge arises due to the incomplete geometry structure and boundary ambiguity of point-cloud instances, leading to inaccurate pseudo labels and poor detection results. To address these challenges, we propose a novel method, named Seg2Box. We first introduce a Multi-Frame Multi-Scale Clustering (MFMS-C) module, which leverages the spatio-temporal consistency of point clouds to generate accurate box-level pseudo-labels. Additionally, the Semantic?Guiding Iterative-Mining Self-Training (SGIM-ST) module is proposed to enhance the performance by progressively refining the pseudo-labels and mining the instances without generating pseudo-labels. Experiments on the Waymo Open Dataset and nuScenes Dataset show that our method significantly outperforms other competitive methods by 23.7\% and 10.3\% in mAP, respectively. The results demonstrate the great label-efficient potential and advancement of our method.
Authors: Yi Niu, Jiashuai Liu, Yingkang Zhan, Jiangbo Shi, Di Zhang, Ines Machado, Mireia Crispin-Ortuzar, Chen Li, Zeyu Gao
Abstract: Spatial Transcriptomics (ST) reveals the spatial distribution of gene expression in tissues, offering critical insights into biological processes and disease mechanisms. However, predicting ST from H\&E-stained histology images is challenging due to the heterogeneous relationship between histomorphology and gene expression, which arises from substantial variability across different patients and tissue sections. A more practical and valuable approach is to utilize ST data from a few local regions to predict the spatial transcriptomic landscape across the remaining regions in H&E slides. In response, we propose PHG2ST, an ST-prompt guided histological hypergraph learning framework, which leverages sparse ST signals as prompts to guide histological hypergraph learning for global spatial gene expression prediction. Our framework fuses histological hypergraph representations at multiple scales through a masked ST-prompt encoding mechanism, improving robustness and generalizability. Benchmark evaluations on two public ST datasets demonstrate that PHG2ST outperforms the existing state-of-the-art methods and closely aligns with the ground truth. These results underscore the potential of leveraging sparse local ST data for scalable and cost-effective spatial gene expression mapping in real-world biomedical applications.
Authors: Yuxin Yao, Zhi Deng, Junhui Hou
Abstract: This paper considers the problem of modeling articulated objects captured in 2D videos to enable novel view synthesis, while also being easily editable, drivable, and re-posable. To tackle this challenging problem, we propose RigGS, a new paradigm that leverages 3D Gaussian representation and skeleton-based motion representation to model dynamic objects without utilizing additional template priors. Specifically, we first propose skeleton-aware node-controlled deformation, which deforms a canonical 3D Gaussian representation over time to initialize the modeling process, producing candidate skeleton nodes that are further simplified into a sparse 3D skeleton according to their motion and semantic information. Subsequently, based on the resulting skeleton, we design learnable skin deformations and pose-dependent detailed deformations, thereby easily deforming the 3D Gaussian representation to generate new actions and render further high-quality images from novel views. Extensive experiments demonstrate that our method can generate realistic new actions easily for objects and achieve high-quality rendering.
Authors: Xiyue Guo, Jiarui Hu, Junjie Hu, Hujun Bao, Guofeng Zhang
Abstract: Recently, camera-based solutions have been extensively explored for scene semantic completion (SSC). Despite their success in visible areas, existing methods struggle to capture complete scene semantics due to frequent visual occlusions. To address this limitation, this paper presents the first satellite-ground cooperative SSC framework, i.e., SGFormer, exploring the potential of satellite-ground image pairs in the SSC task. Specifically, we propose a dual-branch architecture that encodes orthogonal satellite and ground views in parallel, unifying them into a common domain. Additionally, we design a ground-view guidance strategy that corrects satellite image biases during feature encoding, addressing misalignment between satellite and ground views. Moreover, we develop an adaptive weighting strategy that balances contributions from satellite and ground views. Experiments demonstrate that SGFormer outperforms the state of the art on SemanticKITTI and SSCBench-KITTI-360 datasets. Our code is available on https://github.com/gxytcrc/SGFormer.
Authors: Ali Shah Ali, Syed Ahmed Mahmood, Mubin Saeed, Andrey Konin, M. Zeeshan Zia, Quoc-Huy Tran
Abstract: We introduce a novel approach for simultaneous self-supervised video alignment and action segmentation based on a unified optimal transport framework. In particular, we first tackle self-supervised video alignment by developing a fused Gromov-Wasserstein optimal transport formulation with a structural prior, which trains efficiently on GPUs and needs only a few iterations for solving the optimal transport problem. Our single-task method achieves the state-of-the-art performance on multiple video alignment benchmarks and outperforms VAVA, which relies on a traditional Kantorovich optimal transport formulation with an optimality prior. Furthermore, we extend our approach by proposing a unified optimal transport framework for joint self-supervised video alignment and action segmentation, which requires training and storing a single model and saves both time and memory consumption as compared to two different single-task models. Extensive evaluations on several video alignment and action segmentation datasets demonstrate that our multi-task method achieves comparable video alignment yet superior action segmentation results over previous methods in video alignment and action segmentation respectively. Finally, to the best of our knowledge, this is the first work to unify video alignment and action segmentation into a single model.
Authors: Huiqiang Chen, Tianqing Zhu, Linlin Wang, Xin Yu, Longxiang Gao, Wanlei Zhou
Abstract: Large-scale text-to-image (T2I) diffusion models have revolutionized image generation, enabling the synthesis of highly detailed visuals from textual descriptions. However, these models may inadvertently generate inappropriate content, such as copyrighted works or offensive images. While existing methods attempt to eliminate specific unwanted concepts, they often fail to ensure complete removal, allowing the concept to reappear in subtle forms. For instance, a model may successfully avoid generating images in Van Gogh's style when explicitly prompted with 'Van Gogh', yet still reproduce his signature artwork when given the prompt 'Starry Night'. In this paper, we propose SAFER, a novel and efficient approach for thoroughly removing target concepts from diffusion models. At a high level, SAFER is inspired by the observed low-dimensional structure of the text embedding space. The method first identifies a concept-specific subspace $S_c$ associated with the target concept c. It then projects the prompt embeddings onto the complementary subspace of $S_c$, effectively erasing the concept from the generated images. Since concepts can be abstract and difficult to fully capture using natural language alone, we employ textual inversion to learn an optimized embedding of the target concept from a reference image. This enables more precise subspace estimation and enhances removal performance. Furthermore, we introduce a subspace expansion strategy to ensure comprehensive and robust concept erasure. Extensive experiments demonstrate that SAFER consistently and effectively erases unwanted concepts from diffusion models while preserving generation quality.
Authors: Jian Liang, Wenke Huang, Guancheng Wan, Qu Yang, Mang Ye
Abstract: While Multimodal Large Language Models (MLLMs) excel at generalizing across modalities and tasks, effectively adapting them to specific downstream tasks while simultaneously retaining both general and specialized knowledge remains challenging. Although Low-Rank Adaptation (LoRA) is widely used to efficiently acquire specialized knowledge in MLLMs, it introduces substantial harmful redundancy during visual instruction tuning, which exacerbates the forgetting of general knowledge and degrades downstream task performance. To address this issue, we propose LoRASculpt to eliminate harmful redundant parameters, thereby harmonizing general and specialized knowledge. Specifically, under theoretical guarantees, we introduce sparse updates into LoRA to discard redundant parameters effectively. Furthermore, we propose a Conflict Mitigation Regularizer to refine the update trajectory of LoRA, mitigating knowledge conflicts with the pretrained weights. Extensive experimental results demonstrate that even at very high degree of sparsity ($\le$ 5%), our method simultaneously enhances generalization and downstream task performance. This confirms that our approach effectively mitigates the catastrophic forgetting issue and further promotes knowledge harmonization in MLLMs.
Authors: Jiaxi Li, Di Lin, Hao Chen, Hongying Liu, Liang Wan, Wei Feng
Abstract: Deep neural networks (DNNs) often struggle with out-of-distribution data, limiting their reliability in diverse realworld applications. To address this issue, domain generalization methods have been developed to learn domain-invariant features from single or multiple training domains, enabling generalization to unseen testing domains. However, existing approaches usually overlook the impact of style frequency within the training set. This oversight predisposes models to capture spurious visual correlations caused by style confounding factors, rather than learning truly causal representations, thereby undermining inference reliability. In this work, we introduce Style Deconfounding Causal Learning (SDCL), a novel causal inference-based framework designed to explicitly address style as a confounding factor. Our approaches begins with constructing a structural causal model (SCM) tailored to the domain generalization problem and applies a backdoor adjustment strategy to account for style influence. Building on this foundation, we design a style-guided expert module (SGEM) to adaptively clusters style distributions during training, capturing the global confounding style. Additionally, a back-door causal learning module (BDCL) performs causal interventions during feature extraction, ensuring fair integration of global confounding styles into sample predictions, effectively reducing style bias. The SDCL framework is highly versatile and can be seamlessly integrated with state-of-the-art data augmentation techniques. Extensive experiments across diverse natural and medical image recognition tasks validate its efficacy, demonstrating superior performance in both multi-domain and the more challenging single-domain generalization scenarios.
Authors: Zhibo Yang, Wei Hua, Sibo Song, Cong Yao, Yingying Zhu, Wenqing Cheng, Xiang Bai
Abstract: Visual Information Extraction (VIE), aiming at extracting structured information from visually rich document images, plays a pivotal role in document processing. Considering various layouts, semantic scopes, and languages, VIE encompasses an extensive range of types, potentially numbering in the thousands. However, many of these types suffer from a lack of training data, which poses significant challenges. In this paper, we propose a novel generative model, named Generative Compositor, to address the challenge of few-shot VIE. The Generative Compositor is a hybrid pointer-generator network that emulates the operations of a compositor by retrieving words from the source text and assembling them based on the provided prompts. Furthermore, three pre-training strategies are employed to enhance the model's perception of spatial context information. Besides, a prompt-aware resampler is specially designed to enable efficient matching by leveraging the entity-semantic prior contained in prompts. The introduction of the prompt-based retrieval mechanism and the pre-training strategies enable the model to acquire more effective spatial and semantic clues with limited training samples. Experiments demonstrate that the proposed method achieves highly competitive results in the full-sample training, while notably outperforms the baseline in the 1-shot, 5-shot, and 10-shot settings.
Authors: Koki Hirooka, Abu Saleh Musa Miah, Tatsuya Murakami, Yuto Akiba, Yong Seok Hwang, Jungpil Shin
Abstract: Hand gesture-based Sign Language Recognition (SLR) serves as a crucial communication bridge between deaf and non-deaf individuals. Existing SLR systems perform well for their cultural SL but may struggle with multi-cultural sign languages (McSL). To address these challenges, this paper proposes a Stack Spatial-Temporal Transformer Network that leverages multi-head attention mechanisms to capture both spatial and temporal dependencies with hierarchical features using the Stack Transfer concept. In the proceed, firstly, we applied a fully connected layer to make a embedding vector which has high expressive power from the original dataset, then fed them a stack newly proposed transformer to achieve hierarchical features with short-range and long-range dependency. The network architecture is composed of several stages that process spatial and temporal relationships sequentially, ensuring effective feature extraction. After making the fully connected layer, the embedding vector is processed by the Spatial Multi-Head Attention Transformer, which captures spatial dependencies between joints. In the next stage, the Temporal Multi-Head Attention Transformer captures long-range temporal dependencies, and again, the features are concatenated with the output using another skip connection. The processed features are then passed to the Feed-Forward Network (FFN), which refines the feature representations further. After the FFN, additional skip connections are applied to combine the output with earlier layers, followed by a final normalization layer to produce the final output feature tensor. This process is repeated for 10 transformer blocks. The extensive experiment shows that the JSL, KSL and ASL datasets achieved good performance accuracy. Our approach demonstrates improved performance in McSL, and it will be consider as a novel work in this domain.
Authors: Kaisi Guan, Zhengfeng Lai, Yuchong Sun, Peng Zhang, Wei Liu, Kieran Liu, Meng Cao, Ruihua Song
Abstract: Precisely evaluating semantic alignment between text prompts and generated videos remains a challenge in Text-to-Video (T2V) Generation. Existing text-to-video alignment metrics like CLIPScore only generate coarse-grained scores without fine-grained alignment details, failing to align with human preference. To address this limitation, we propose ETVA, a novel Evaluation method of Text-to-Video Alignment via fine-grained question generation and answering. First, a multi-agent system parses prompts into semantic scene graphs to generate atomic questions. Then we design a knowledge-augmented multi-stage reasoning framework for question answering, where an auxiliary LLM first retrieves relevant common-sense knowledge (e.g., physical laws), and then video LLM answers the generated questions through a multi-stage reasoning mechanism. Extensive experiments demonstrate that ETVA achieves a Spearman's correlation coefficient of 58.47, showing a much higher correlation with human judgment than existing metrics which attain only 31.0. We also construct a comprehensive benchmark specifically designed for text-to-video alignment evaluation, featuring 2k diverse prompts and 12k atomic questions spanning 10 categories. Through a systematic evaluation of 15 existing text-to-video models, we identify their key capabilities and limitations, paving the way for next-generation T2V generation.
Authors: Dongseob Kim, Hyunjung Shim
Abstract: Multi-label classification is crucial for comprehensive image understanding, yet acquiring accurate annotations is challenging and costly. To address this, a recent study suggests exploiting unsupervised multi-label classification leveraging CLIP, a powerful vision-language model. Despite CLIP's proficiency, it suffers from view-dependent predictions and inherent bias, limiting its effectiveness. We propose a novel method that addresses these issues by leveraging multiple views near target objects, guided by Class Activation Mapping (CAM) of the classifier, and debiasing pseudo-labels derived from CLIP predictions. Our Classifier-guided CLIP Distillation (CCD) enables selecting multiple local views without extra labels and debiasing predictions to enhance classification performance. Experimental results validate our method's superiority over existing techniques across diverse datasets. The code is available at https://github.com/k0u-id/CCD.
Authors: Yu Qiu, Yuhang Sun, Jie Mei, Lin Xiao, Jing Xu
Abstract: Traffic Salient Object Detection (TSOD) aims to segment the objects critical to driving safety by combining semantic (e.g., collision risks) and visual saliency. Unlike SOD in natural scene images (NSI-SOD), which prioritizes visually distinctive regions, TSOD emphasizes the objects that demand immediate driver attention due to their semantic impact, even with low visual contrast. This dual criterion, i.e., bridging perception and contextual risk, re-defines saliency for autonomous and assisted driving systems. To address the lack of task-specific benchmarks, we collect the first large-scale TSOD dataset with pixel-wise saliency annotations, named TSOD10K. TSOD10K covers the diverse object categories in various real-world traffic scenes under various challenging weather/illumination variations (e.g., fog, snowstorms, low-contrast, and low-light). Methodologically, we propose a Mamba-based TSOD model, termed Tramba. Considering the challenge of distinguishing inconspicuous visual information from complex traffic backgrounds, Tramba introduces a novel Dual-Frequency Visual State Space module equipped with shifted window partitioning and dilated scanning to enhance the perception of fine details and global structure by hierarchically decomposing high/low-frequency components. To emphasize critical regions in traffic scenes, we propose a traffic-oriented Helix 2D-Selective-Scan (Helix-SS2D) mechanism that injects driving attention priors while effectively capturing global multi-direction spatial dependencies. We establish a comprehensive benchmark by evaluating Tramba and 22 existing NSI-SOD models on TSOD10K, demonstrating Tramba's superiority. Our research establishes the first foundation for safety-aware saliency analysis in intelligent transportation systems.
Authors: Xiaoyong Chen, Yong Guo, Jiaming Liang, Sitong Zhuang, Runhao Zeng, Xiping Hu
Abstract: Temporal action detection (TAD) aims to identify and localize action instances in untrimmed videos, which is essential for various video understanding tasks. However, recent improvements in model performance, driven by larger feature extractors and datasets, have led to increased computational demands. This presents a challenge for applications like autonomous driving and robotics, which rely on limited computational resources. While existing channel pruning methods can compress these models, reducing the number of channels often hinders the parallelization efficiency of GPU, due to the inefficient multiplication between small matrices. Instead of pruning channels, we propose a Progressive Block Drop method that reduces model depth while retaining layer width. In this way, we still use large matrices for computation but reduce the number of multiplications. Our approach iteratively removes redundant blocks in two steps: first, we drop blocks with minimal impact on model performance; and second, we employ a parameter-efficient cross-depth alignment technique, fine-tuning the pruned model to restore model accuracy. Our method achieves a 25% reduction in computational overhead on two TAD benchmarks (THUMOS14 and ActivityNet-1.3) to achieve lossless compression. More critically, we empirically show that our method is orthogonal to channel pruning methods and can be combined with it to yield further efficiency gains.
Authors: Lingfan Zhang, Chen Liu, Chengming Xu, Kai Hu, Donghao Luo, Chengjie Wang, Yanwei Fu, Yuan Yao
Abstract: In recent years, the field of image generation has witnessed significant advancements, particularly in fine-tuning methods that align models with universal human preferences. This paper explores the critical role of preference data in the training process of diffusion models, particularly in the context of Diffusion-DPO and its subsequent adaptations. We investigate the complexities surrounding universal human preferences in image generation, highlighting the subjective nature of these preferences and the challenges posed by minority samples in preference datasets. Through pilot experiments, we demonstrate the existence of minority samples and their detrimental effects on model performance. We propose Adaptive-DPO -- a novel approach that incorporates a minority-instance-aware metric into the DPO objective. This metric, which includes intra-annotator confidence and inter-annotator stability, distinguishes between majority and minority samples. We introduce an Adaptive-DPO loss function which improves the DPO loss in two ways: enhancing the model's learning of majority labels while mitigating the negative impact of minority samples. Our experiments demonstrate that this method effectively handles both synthetic minority data and real-world preference data, paving the way for more effective training methodologies in image generation tasks.
Authors: Joo Chan Lee, Jong Hwan Ko, Eunbyung Park
Abstract: 3D Gaussian Splatting (3DGS) has emerged as a powerful representation for real-time, high-performance rendering, enabling a wide range of applications. However, representing 3D scenes with numerous explicit Gaussian primitives imposes significant storage and memory overhead. Recent studies have shown that high-quality rendering can be achieved with a substantially reduced number of Gaussians when represented with high-precision attributes. Nevertheless, existing 3DGS compression methods still rely on a relatively large number of Gaussians, focusing primarily on attribute compression. This is because a smaller set of Gaussians becomes increasingly sensitive to lossy attribute compression, leading to severe quality degradation. Since the number of Gaussians is directly tied to computational costs, it is essential to reduce the number of Gaussians effectively rather than only optimizing storage. In this paper, we propose Optimized Minimal Gaussians representation (OMG), which significantly reduces storage while using a minimal number of primitives. First, we determine the distinct Gaussian from the near ones, minimizing redundancy without sacrificing quality. Second, we propose a compact and precise attribute representation that efficiently captures both continuity and irregularity among primitives. Additionally, we propose a sub-vector quantization technique for improved irregularity representation, maintaining fast training with a negligible codebook size. Extensive experiments demonstrate that OMG reduces storage requirements by nearly 50% compared to the previous state-of-the-art and enables 600+ FPS rendering while maintaining high rendering quality. Our source code is available at https://maincold2.github.io/omg/.
Authors: Shicheng Li, Lei Li, Kun Ouyang, Shuhuai Ren, Yuanxin Liu, Yuanxing Zhang, Fuzheng Zhang, Lingpeng Kong, Qi Liu, Xu Sun
Abstract: Video Large Language Models (Video LLMs) have achieved significant success by leveraging a two-stage paradigm: pretraining on large-scale video-text data for vision-language alignment, followed by supervised fine-tuning (SFT) for task-specific capabilities. However, existing approaches struggle with temporal reasoning due to weak temporal correspondence in the data and reliance on the next-token prediction paradigm during training. To address these limitations, we propose TEMPO (TEMporal Preference Optimization), a systematic framework that enhances Video LLMs' temporal reasoning capabilities through Direct Preference Optimization (DPO). To facilitate this, we introduce an automated preference data generation pipeline that systematically constructs preference pairs by selecting videos that are rich in temporal information, designing video-specific perturbation strategies, and finally evaluating model responses on clean and perturbed video inputs. Our temporal alignment features two key innovations: curriculum learning which that progressively increases perturbation difficulty to improve model robustness and adaptability; and ``Pre-SFT Alignment'', applying preference optimization before instruction tuning to prioritize fine-grained temporal comprehension. Extensive experiments demonstrate that our approach consistently improves Video LLM performance across multiple benchmarks with a relatively small set of self-generated DPO data. We further analyze the transferability of DPO data across architectures and the role of difficulty scheduling in optimization. Our findings highlight our TEMPO as a scalable and efficient complement to SFT-based methods, paving the way for developing reliable Video LLMs.
Authors: Haijin Zeng, Xiangming Wang, Yongyong Chen, Jingyong Su, Jie Liu
Abstract: Dynamic image degradations, including noise, blur and lighting inconsistencies, pose significant challenges in image restoration, often due to sensor limitations or adverse environmental conditions. Existing Deep Unfolding Networks (DUNs) offer stable restoration performance but require manual selection of degradation matrices for each degradation type, limiting their adaptability across diverse scenarios. To address this issue, we propose the Vision-Language-guided Unfolding Network (VLU-Net), a unified DUN framework for handling multiple degradation types simultaneously. VLU-Net leverages a Vision-Language Model (VLM) refined on degraded image-text pairs to align image features with degradation descriptions, selecting the appropriate transform for target degradation. By integrating an automatic VLM-based gradient estimation strategy into the Proximal Gradient Descent (PGD) algorithm, VLU-Net effectively tackles complex multi-degradation restoration tasks while maintaining interpretability. Furthermore, we design a hierarchical feature unfolding structure to enhance VLU-Net framework, efficiently synthesizing degradation patterns across various levels. VLU-Net is the first all-in-one DUN framework and outperforms current leading one-by-one and all-in-one end-to-end methods by 3.74 dB on the SOTS dehazing dataset and 1.70 dB on the Rain100L deraining dataset.
Authors: Yingying Fan, Quanwei Yang, Kaisiyuan Wang, Hang Zhou, Yingying Li, Haocheng Feng, Yu Wu, Jingdong Wang
Abstract: Current digital human studies focusing on lip-syncing and body movement are no longer sufficient to meet the growing industrial demand, while human video generation techniques that support interacting with real-world environments (e.g., objects) have not been well investigated. Despite human hand synthesis already being an intricate problem, generating objects in contact with hands and their interactions presents an even more challenging task, especially when the objects exhibit obvious variations in size and shape. To cope with these issues, we present a novel video Reenactment framework focusing on Human-Object Interaction (HOI) via an adaptive Layout-instructed Diffusion model (Re-HOLD). Our key insight is to employ specialized layout representation for hands and objects, respectively. Such representations enable effective disentanglement of hand modeling and object adaptation to diverse motion sequences. To further improve the generation quality of HOI, we have designed an interactive textural enhancement module for both hands and objects by introducing two independent memory banks. We also propose a layout-adjusting strategy for the cross-object reenactment scenario to adaptively adjust unreasonable layouts caused by diverse object sizes during inference. Comprehensive qualitative and quantitative evaluations demonstrate that our proposed framework significantly outperforms existing methods. Project page: https://fyycs.github.io/Re-HOLD.
Authors: Mengtian Li, Jinshu Chen, Wanquan Feng, Bingchuan Li, Fei Dai, Songtao Zhao, Qian He
Abstract: Personalized portrait synthesis, essential in domains like social entertainment, has recently made significant progress. Person-wise fine-tuning based methods, such as LoRA and DreamBooth, can produce photorealistic outputs but need training on individual samples, consuming time and resources and posing an unstable risk. Adapter based techniques such as IP-Adapter freeze the foundational model parameters and employ a plug-in architecture to enable zero-shot inference, but they often exhibit a lack of naturalness and authenticity, which are not to be overlooked in portrait synthesis tasks. In this paper, we introduce a parameter-efficient adaptive generation method, namely HyperLoRA, that uses an adaptive plug-in network to generate LoRA weights, merging the superior performance of LoRA with the zero-shot capability of adapter scheme. Through our carefully designed network structure and training strategy, we achieve zero-shot personalized portrait generation (supporting both single and multiple image inputs) with high photorealism, fidelity, and editability.
Authors: Ibtissam Saadi, Abdenour Hadid, Douglas W. Cunningham, Abdelmalik Taleb-Ahmed, Yassin El Hillali
Abstract: Vision-Language Models (VLMs) like CLIP offer promising solutions for Dynamic Facial Expression Recognition (DFER) but face challenges such as inefficient full fine-tuning, high complexity, and poor alignment between textual and visual representations. Additionally, existing methods struggle with ineffective temporal modeling. To address these issues, we propose PE-CLIP, a parameter-efficient fine-tuning (PEFT) framework that adapts CLIP for DFER while significantly reducing trainable parameters while maintaining high accuracy. PE-CLIP introduces two specialized adapters: a Temporal Dynamic Adapter (TDA) and a Shared Adapter (ShA). The TDA is a GRU-based module with dynamic scaling that captures sequential dependencies while emphasizing informative temporal features and suppressing irrelevant variations. The ShA is a lightweight adapter that refines representations within both textual and visual encoders, ensuring consistency and efficiency. Additionally, we integrate Multi-modal Prompt Learning (MaPLe), introducing learnable prompts for visual and action unit-based textual inputs, enhancing semantic alignment between modalities and enabling efficient CLIP adaptation for dynamic tasks. We evaluate PE-CLIP on two benchmark datasets, DFEW and FERV39K, achieving competitive performance compared to state-of-the-art methods while requiring fewer trainable parameters. By balancing efficiency and accuracy, PE-CLIP sets a new benchmark in resource-efficient DFER. The source code of the proposed PE-CLIP will be publicly available at https://github.com/Ibtissam-SAADI/PE-CLIP .
Authors: Yinhan Zhang, Yue Ma, Bingyuan Wang, Qifeng Chen, Zeyu Wang
Abstract: We present \textit{MagicColor}, a diffusion-based framework for multi-instance sketch colorization. The production of multi-instance 2D line art colorization adheres to an industry-standard workflow, which consists of three crucial stages: the design of line art characters, the coloring of individual objects, and the refinement process. The artists are required to repeat the process of coloring each instance one by one, which is inaccurate and inefficient. Meanwhile, current generative methods fail to solve this task due to the challenge of multi-instance pair data collection. To tackle these challenges, we incorporate three technical designs to ensure precise character detail transcription and achieve multi-instance sketch colorization in a single forward. Specifically, we first propose the self-play training strategy to solve the lack of training data. Then we introduce an instance guider to feed the color of the instance. To achieve accurate color matching, we present fine-grained color matching with edge loss to enhance visual quality. Equipped with the proposed modules, MagicColor enables automatically transforming sketches into vividly-colored images with accurate consistency and multi-instance control. Experiments on our collected datasets show that our model outperforms existing methods regarding chromatic precision. Specifically, our model critically automates the colorization process with zero manual adjustments, so novice users can produce stylistically consistent artwork by providing reference instances and the original line art. Our code and additional details are available at https://yinhan-zhang.github.io/color
Authors: Wei Zhang, Mengting Ma, Yizhen Jiang, Rongrong Lian, Zhenkai Wu, Kangning Cui, Xiaowen Ma
Abstract: Compared with natural images, remote sensing images (RSIs) have the unique characteristic. i.e., larger intraclass variance, which makes semantic segmentation for remote sensing images more challenging. Moreover, existing semantic segmentation models for remote sensing images usually employ a vanilla softmax classifier, which has three drawbacks: (1) non-direct supervision for the pixel representations during training; (2) inadequate modeling ability of parametric softmax classifiers under large intraclass variance; and (3) opaque process of classification decision. In this paper, we propose a novel classifier (called CenterSeg) customized for RSI semantic segmentation, which solves the abovementioned problems with multiple prototypes, direct supervision under Grassmann manifold, and interpretability strategy. Specifically, for each class, our CenterSeg obtains local class centers by aggregating corresponding pixel features based on ground-truth masks, and generates multiple prototypes through hard attention assignment and momentum updating. In addition, we introduce the Grassmann manifold and constrain the joint embedding space of pixel features and prototypes based on two additional regularization terms. Especially, during the inference, CenterSeg can further provide interpretability to the model by restricting the prototype as a sample of the training set. Experimental results on three remote sensing segmentation datasets validate the effectiveness of the model. Besides the superior performance, CenterSeg has the advantages of simplicity, lightweight, compatibility, and interpretability. Code is available at https://github.com/xwmaxwma/rssegmentation.
Authors: Jiadong Tang, Yu Gao, Dianyi Yang, Liqi Yan, Yufeng Yue, Yi Yang
Abstract: Drones have become essential tools for reconstructing wild scenes due to their outstanding maneuverability. Recent advances in radiance field methods have achieved remarkable rendering quality, providing a new avenue for 3D reconstruction from drone imagery. However, dynamic distractors in wild environments challenge the static scene assumption in radiance fields, while limited view constraints hinder the accurate capture of underlying scene geometry. To address these challenges, we introduce DroneSplat, a novel framework designed for robust 3D reconstruction from in-the-wild drone imagery. Our method adaptively adjusts masking thresholds by integrating local-global segmentation heuristics with statistical approaches, enabling precise identification and elimination of dynamic distractors in static scenes. We enhance 3D Gaussian Splatting with multi-view stereo predictions and a voxel-guided optimization strategy, supporting high-quality rendering under limited view constraints. For comprehensive evaluation, we provide a drone-captured 3D reconstruction dataset encompassing both dynamic and static scenes. Extensive experiments demonstrate that DroneSplat outperforms both 3DGS and NeRF baselines in handling in-the-wild drone imagery.
Authors: Yingping Liang, Yutao Hu, Wenqi Shao, Ying Fu
Abstract: Depth completion involves predicting dense depth maps from sparse LiDAR inputs. However, sparse depth annotations from sensors limit the availability of dense supervision, which is necessary for learning detailed geometric features. In this paper, we propose a two-stage knowledge distillation framework that leverages powerful monocular foundation models to provide dense supervision for depth completion. In the first stage, we introduce a pre-training strategy that generates diverse training data from natural images, which distills geometric knowledge to depth completion. Specifically, we simulate LiDAR scans by utilizing monocular depth and mesh reconstruction, thereby creating training data without requiring ground-truth depth. Besides, monocular depth estimation suffers from inherent scale ambiguity in real-world settings. To address this, in the second stage, we employ a scale- and shift-invariant loss (SSI Loss) to learn real-world scales when fine-tuning on real-world datasets. Our two-stage distillation framework enables depth completion models to harness the strengths of monocular foundation models. Experimental results demonstrate that models trained with our two-stage distillation framework achieve state-of-the-art performance, ranking \textbf{first place} on the KITTI benchmark. Code is available at https://github.com/Sharpiless/DMD3C
Authors: Wentao Jiang, Jingya Wang, Haotao Lu, Kaiyang Ji, Baoxiong Jia, Siyuan Huang, Ye Shi
Abstract: Human action-reaction synthesis, a fundamental challenge in modeling causal human interactions, plays a critical role in applications ranging from virtual reality to social robotics. While diffusion-based models have demonstrated promising performance, they exhibit two key limitations for interaction synthesis: reliance on complex noise-to-reaction generators with intricate conditional mechanisms, and frequent physical violations in generated motions. To address these issues, we propose Action-Reaction Flow Matching (ARFlow), a novel framework that establishes direct action-to-reaction mappings, eliminating the need for complex conditional mechanisms. Our approach introduces two key innovations: an x1-prediction method that directly outputs human motions instead of velocity fields, enabling explicit constraint enforcement; and a training-free, gradient-based physical guidance mechanism that effectively prevents body penetration artifacts during sampling. Extensive experiments on NTU120 and Chi3D datasets demonstrate that ARFlow not only outperforms existing methods in terms of Fr\'echet Inception Distance and motion diversity but also significantly reduces body collisions, as measured by our new Intersection Volume and Intersection Frequency metrics.
Authors: Xiaofeng Mao, Yuefeng Chen, Rong Zhang, Hui Xue, Zhao Li, Hang Su
Abstract: Deep neural networks (DNNs) has shown great promise in computer vision tasks. However, machine vision achieved by DNNs cannot be as robust as human perception. Adversarial attacks and data distribution shifts have been known as two major scenarios which degrade machine performance and obstacle the wide deployment of machines "in the wild". In order to break these obstructions and facilitate the research of model robustness, we develop EasyRobust, a comprehensive and easy-to-use toolkit for training, evaluation and analysis of robust vision models. EasyRobust targets at two types of robustness: 1) Adversarial robustness enables the model to defense against malicious inputs crafted by worst-case perturbations, also known as adversarial examples; 2) Non-adversarial robustness enhances the model performance on natural test images with corruptions or distribution shifts. Thorough benchmarks on image classification enable EasyRobust to provide an accurate robustness evaluation on vision models. We wish our EasyRobust can help for training practically-robust models and promote academic and industrial progress in closing the gap between human and machine vision. Codes and models of EasyRobust have been open-sourced in https://github.com/alibaba/easyrobust.
Authors: Weihao Yu, Xiaoqing Guo, Chenxin Li, Yifan Liu, Yixuan Yuan
Abstract: Achieving meticulous segmentation of tooth point clouds from intra-oral scans stands as an indispensable prerequisite for various orthodontic applications. Given the labor-intensive nature of dental annotation, a significant amount of data remains unlabeled, driving increasing interest in semi-supervised approaches. One primary challenge of existing semi-supervised medical segmentation methods lies in noisy pseudo labels generated for unlabeled data. To address this challenge, we propose GeoT, the first framework that employs instance-dependent transition matrix (IDTM) to explicitly model noise in pseudo labels for semi-supervised dental segmentation. Specifically, to handle the extensive solution space of IDTM arising from tens of thousands of dental points, we introduce tooth geometric priors through two key components: point-level geometric regularization (PLGR) to enhance consistency between point adjacency relationships in 3D and IDTM spaces, and class-level geometric smoothing (CLGS) to leverage the fixed spatial distribution of tooth categories for optimal IDTM estimation. Extensive experiments performed on the public Teeth3DS dataset and private dataset demonstrate that our method can make full utilization of unlabeled data to facilitate segmentation, achieving performance comparable to fully supervised methods with only $20\%$ of the labeled data.
Authors: Jinbo Yan, Rui Peng, Zhiyan Wang, Luyang Tang, Jiayu Yang, Jie Liang, Jiahao Wu, Ronggang Wang
Abstract: Building Free-Viewpoint Videos in a streaming manner offers the advantage of rapid responsiveness compared to offline training methods, greatly enhancing user experience. However, current streaming approaches face challenges of high per-frame reconstruction time (10s+) and error accumulation, limiting their broader application. In this paper, we propose Instant Gaussian Stream (IGS), a fast and generalizable streaming framework, to address these issues. First, we introduce a generalized Anchor-driven Gaussian Motion Network, which projects multi-view 2D motion features into 3D space, using anchor points to drive the motion of all Gaussians. This generalized Network generates the motion of Gaussians for each target frame in the time required for a single inference. Second, we propose a Key-frame-guided Streaming Strategy that refines each key frame, enabling accurate reconstruction of temporally complex scenes while mitigating error accumulation. We conducted extensive in-domain and cross-domain evaluations, demonstrating that our approach can achieve streaming with a average per-frame reconstruction time of 2s+, alongside a enhancement in view synthesis quality.
Authors: Haichao Zhang, Zhuowei Li, Dimitris Metaxas, Yun Fu
Abstract: Token-based video representation has emerged as a promising approach for enabling large language models to interpret video content. However, existing token reduction techniques, such as token pruning and token merging, often disrupt essential spatial-temporal positional embeddings, failing to adequately balance computational efficiency with fewer tokens. Consequently, these methods result in relatively lengthy token sequences, limiting their applicability in scenarios requiring extreme token compression, such as video large language models. In this paper, we introduce the novel task of extreme short token reduction, aiming to represent extensive video sequences with a minimal number of tokens. To address this challenge, we propose Token Dynamics, a new video representation framework that dynamically reduces token count while preserving spatial-temporal coherence. Specifically, we disentangle video representations by separating visual embeddings from grid-level motion information, structuring them into: 1. a concise token base, created by clustering tokens that describe object-level content; 2. a token dynamics map, capturing detailed spatial-temporal motion patterns across grids. Furthermore, we introduce a cross-dynamics attention mechanism that integrates motion features into the token base without increasing token length, thereby maintaining compactness and spatial-temporal integrity. The experiments demonstrate a reduction of token count to merely 0.07% of the original tokens, with only a minor performance drop of 1.13%. Additionally, we propose two novel subtasks within extreme token reduction (fixed-length and adaptive-length compression), both effectively representing long token sequences for video-language tasks. Our method offers significantly lower theoretical complexity, fewer tokens, and enhanced throughput, thus providing an efficient solution for video LLMs.
Authors: Xu Zhang, Hao Zhou, Haoming Qin, Xiaobin Lu, Jiaxing Yan, Guanzhong Wang, Zeyu Chen, Yi Liu
Abstract: Despite substantial progress in text-to-video generation, achieving precise and flexible control over fine-grained spatiotemporal attributes remains a significant unresolved challenge in video generation research. To address these limitations, we introduce VCtrl (also termed PP-VCtrl), a novel framework designed to enable fine-grained control over pre-trained video diffusion models in a unified manner. VCtrl integrates diverse user-specified control signals-such as Canny edges, segmentation masks, and human keypoints-into pretrained video diffusion models via a generalizable conditional module capable of uniformly encoding multiple types of auxiliary signals without modifying the underlying generator. Additionally, we design a unified control signal encoding pipeline and a sparse residual connection mechanism to efficiently incorporate control representations. Comprehensive experiments and human evaluations demonstrate that VCtrl effectively enhances controllability and generation quality. The source code and pre-trained models are publicly available and implemented using the PaddlePaddle framework at http://github.com/PaddlePaddle/PaddleMIX/tree/develop/ppdiffusers/examples/ppvctrl.
URLs: http://github.com/PaddlePaddle/PaddleMIX/tree/develop/ppdiffusers/examples/ppvctrl.
Authors: Qinghe Ma, Jian Zhang, Zekun Li, Lei Qi, Qian Yu, Yinghuan Shi
Abstract: Large pretrained visual foundation models exhibit impressive general capabilities. However, the extensive prior knowledge inherent in these models can sometimes be a double-edged sword when adapting them to downstream tasks in specific domains. In the context of semi-supervised medical image segmentation with domain shift, foundation models like MedSAM tend to make overconfident predictions, some of which are incorrect. The error accumulation hinders the effective utilization of unlabeled data and limits further improvements. In this paper, we introduce a Synergistic training framework for Foundation and Conventional models (SynFoC) to address the issue. We observe that a conventional model trained from scratch has the ability to correct the high-confidence mispredictions of the foundation model, while the foundation model can supervise it with high-quality pseudo-labels in the early training stages. Furthermore, to enhance the collaborative training effectiveness of both models and promote reliable convergence towards optimization, the consensus-divergence consistency regularization is proposed. We demonstrate the superiority of our method across four public multi-domain datasets. In particular, our method improves the Dice score by 10.31\% on the Prostate dataset. Our code is available at https://github.com/MQinghe/SynFoC .
Authors: Ziteng Cui, Jianfei Yang, Tatsuya Harada
Abstract: In the computer vision community, the preference for pre-training visual models has largely shifted toward sRGB images due to their ease of acquisition and compact storage. However, camera RAW images preserve abundant physical details across diverse real-world scenarios. Despite this, most existing visual perception methods that utilize RAW data directly integrate image signal processing (ISP) stages with subsequent network modules, often overlooking potential synergies at the model level. Building on recent advances in adapter-based methodologies in both NLP and computer vision, we propose RAW-Adapter, a novel framework that incorporates learnable ISP modules as input-level adapters to adjust RAW inputs. At the same time, it employs model-level adapters to seamlessly bridge ISP processing with high-level downstream architectures. Moreover, RAW-Adapter serves as a general framework applicable to various computer vision frameworks. Furthermore, we introduce RAW-Bench, which incorporates 17 types of RAW-based common corruptions, including lightness degradations, weather effects, blurriness, camera imaging degradations, and variations in camera color response. Using this benchmark, we systematically compare the performance of RAW-Adapter with state-of-the-art (SOTA) ISP methods and other RAW-based high-level vision algorithms. Additionally, we propose a RAW-based data augmentation strategy to further enhance RAW-Adapter's performance and improve its out-of-domain (OOD) generalization ability. Extensive experiments substantiate the effectiveness and efficiency of RAW-Adapter, highlighting its robust performance across diverse scenarios.
Authors: Junjie Hu, Shuyong Gao, Qianyu Guo, Yan Wang, Qishan Wang, Yuang Feng, Wenqiang Zhang
Abstract: Humans can intuitively decompose an image into a sequence of strokes to create a painting, yet existing methods for generating drawing processes are limited to specific data types and often rely on expensive human-annotated datasets. We propose a novel self-supervised framework for generating drawing processes from any type of image, treating the task as a video generation problem. Our approach reverses the drawing process by progressively removing strokes from a reference image, simulating a human-like creation sequence. Crucially, our method does not require costly datasets of real human drawing processes; instead, we leverage depth estimation and stroke rendering to construct a self-supervised dataset. We model human drawings as "refinement" and "layering" processes and introduce depth fusion layers to enable video generation models to learn and replicate human drawing behavior. Extensive experiments validate the effectiveness of our approach, demonstrating its ability to generate realistic drawings without the need for real drawing process data.
Authors: Jianchuan Chen, Jingchuan Hu, Gaige Wang, Zhonghua Jiang, Tiansong Zhou, Zhiwen Chen, Chengfei Lv
Abstract: Realistic 3D full-body talking avatars hold great potential in AR, with applications ranging from e-commerce live streaming to holographic communication. Despite advances in 3D Gaussian Splatting (3DGS) for lifelike avatar creation, existing methods struggle with fine-grained control of facial expressions and body movements in full-body talking tasks. Additionally, they often lack sufficient details and cannot run in real-time on mobile devices. We present TaoAvatar, a high-fidelity, lightweight, 3DGS-based full-body talking avatar driven by various signals. Our approach starts by creating a personalized clothed human parametric template that binds Gaussians to represent appearances. We then pre-train a StyleUnet-based network to handle complex pose-dependent non-rigid deformation, which can capture high-frequency appearance details but is too resource-intensive for mobile devices. To overcome this, we "bake" the non-rigid deformations into a lightweight MLP-based network using a distillation technique and develop blend shapes to compensate for details. Extensive experiments show that TaoAvatar achieves state-of-the-art rendering quality while running in real-time across various devices, maintaining 90 FPS on high-definition stereo devices such as the Apple Vision Pro.
Authors: Stephen Lloyd-Brown, Susan Francis, Caroline Hoad, Penny Gowland, Karen Mullinger, Andrew French, Xin Chen
Abstract: An often overlooked problem in medical image segmentation research is the effective selection of training subsets to annotate from a complete set of unlabelled data. Many studies select their training sets at random, which may lead to suboptimal model performance, especially in the minimal supervision setting where each training image has a profound effect on performance outcomes. This work aims to address this issue. We use prototypical contrasting learning and clustering to extract representative and diverse samples for annotation. We improve upon prior works with a bespoke cluster-based image selection process. Additionally, we introduce the concept of unsupervised balanced batch dataloading to medical image segmentation, which aims to improve model learning with minimally annotated data. We evaluated our method on a public skin lesion dataset (ISIC 2018) and compared it to another state-of-the-art data sampling method. Our method achieved superior performance in a low annotation budget scenario.
Authors: Chandan Yeshwanth, David Rozenberszki, Angela Dai
Abstract: Generating text descriptions of objects in 3D indoor scenes is an important building block of embodied understanding. Existing methods do this by describing objects at a single level of detail, which often does not capture fine-grained details such as varying textures, materials, and shapes of the parts of objects. We propose the task of expressive 3D captioning: given an input 3D scene, describe objects at multiple levels of detail: a high-level object description, and a low-level description of the properties of its parts. To produce such captions, we present ExCap3D, an expressive 3D captioning model which takes as input a 3D scan, and for each detected object in the scan, generates a fine-grained collective description of the parts of the object, along with an object-level description conditioned on the part-level description. We design ExCap3D to encourage semantic consistency between the generated text descriptions, as well as textual similarity in the latent space, to further increase the quality of the generated captions. To enable this task, we generated the ExCap3D Dataset by leveraging a visual-language model (VLM) for multi-view captioning. The ExCap3D Dataset contains captions on the ScanNet++ dataset with varying levels of detail, comprising 190k text descriptions of 34k 3D objects in 947 indoor scenes. Our experiments show that the object- and part-level of detail captions generated by ExCap3D are of higher quality than those produced by state-of-the-art methods, with a Cider score improvement of 17% and 124% for object- and part-level details respectively. Our code, dataset and models will be made publicly available.
Authors: Yuang Feng, Shuyong Gao, Fuzhen Yan, Yicheng Song, Lingyi Hong, Junjie Hu, Wenqiang Zhang
Abstract: Video Camouflaged Object Detection (VCOD) aims to segment objects whose appearances closely resemble their surroundings, posing a challenging and emerging task. Existing vision models often struggle in such scenarios due to the indistinguishable appearance of camouflaged objects and the insufficient exploitation of dynamic information in videos. To address these challenges, we propose an end-to-end VCOD framework inspired by human memory-recognition, which leverages historical video information by integrating memory reference frames for camouflaged sequence processing. Specifically, we design a dual-purpose decoder that simultaneously generates predicted masks and scores, enabling reference frame selection based on scores while introducing auxiliary supervision to enhance feature extraction.Furthermore, this study introduces a novel reference-guided multilevel asymmetric attention mechanism, effectively integrating long-term reference information with short-term motion cues for comprehensive feature extraction. By combining these modules, we develop the Scoring, Remember, and Reference (SRR) framework, which efficiently extracts information to locate targets and employs memory guidance to improve subsequent processing. With its optimized module design and effective utilization of video data, our model achieves significant performance improvements, surpassing existing approaches by 10% on benchmark datasets while requiring fewer parameters (54M) and only a single pass through the video. The code will be made publicly available.
Authors: Yufei Shi, Weilong Yan, Gang Xu, Yumeng Li, Yuchen Li, Zhenxi Li, Fei Richard Yu, Ming Li, Si Yong Yeo
Abstract: Video large language models (ViLLMs) excel in general video understanding, e.g., recognizing activities like talking and eating, but struggle with identity-aware comprehension, such as "Wilson is receiving chemotherapy" or "Tom is discussing with Sarah", limiting their applicability in smart healthcare and smart home environments. To address this limitation, we propose a one-shot learning framework PVChat, the first personalized ViLLM that enables subject-aware question answering (QA) from a single video for each subject. Our approach optimizes a Mixture-of-Heads (MoH) enhanced ViLLM on a synthetically augmented video-QA dataset, leveraging a progressive image-to-video learning strategy. Specifically, we introduce an automated augmentation pipeline that synthesizes identity-preserving positive samples and retrieves hard negatives from existing video corpora, generating a diverse training dataset with four QA types: existence, appearance, action, and location inquiries. To enhance subject-specific learning, we propose a ReLU Routing MoH attention mechanism, alongside two novel objectives: (1) Smooth Proximity Regularization for progressive learning through exponential distance scaling and (2) Head Activation Enhancement for balanced attention routing. Finally, we adopt a two-stage training strategy, transitioning from image pre-training to video fine-tuning, enabling a gradual learning process from static attributes to dynamic representations. We evaluate PVChat on diverse datasets covering medical scenarios, TV series, anime, and real-world footage, demonstrating its superiority in personalized feature understanding after learning from a single video, compared to state-of-the-art ViLLMs.
Authors: Pablo Garcia-Fernandez, Lorenzo Vaquero, Mingxuan Liu, Feng Xue, Daniel Cores, Nicu Sebe, Manuel Mucientes, Elisa Ricci
Abstract: Open-vocabulary object detection (OvOD) is set to revolutionize security screening by enabling systems to recognize any item in X-ray scans. However, developing effective OvOD models for X-ray imaging presents unique challenges due to data scarcity and the modality gap that prevents direct adoption of RGB-based solutions. To overcome these limitations, we propose RAXO, a training-free framework that repurposes off-the-shelf RGB OvOD detectors for robust X-ray detection. RAXO builds high-quality X-ray class descriptors using a dual-source retrieval strategy. It gathers relevant RGB images from the web and enriches them via a novel X-ray material transfer mechanism, eliminating the need for labeled databases. These visual descriptors replace text-based classification in OvOD, leveraging intra-modal feature distances for robust detection. Extensive experiments demonstrate that RAXO consistently improves OvOD performance, providing an average mAP increase of up to 17.0 points over base detectors. To further support research in this emerging field, we also introduce DET-COMPASS, a new benchmark featuring bounding box annotations for over 300 object categories, enabling large-scale evaluation of OvOD in X-ray. Code and dataset available at: https://github.com/PAGF188/RAXO.
Authors: Vittorio Pippi, Fabio Quattrini, Silvia Cascianelli, Alessio Tonioni, Rita Cucchiara
Abstract: Styled Handwritten Text Generation (HTG) has recently received attention from the computer vision and document analysis communities, which have developed several solutions, either GAN- or diffusion-based, that achieved promising results. Nonetheless, these strategies fail to generalize to novel styles and have technical constraints, particularly in terms of maximum output length and training efficiency. To overcome these limitations, in this work, we propose a novel framework for text image generation, dubbed Emuru. Our approach leverages a powerful text image representation model (a variational autoencoder) combined with an autoregressive Transformer. Our approach enables the generation of styled text images conditioned on textual content and style examples, such as specific fonts or handwriting styles. We train our model solely on a diverse, synthetic dataset of English text rendered in over 100,000 typewritten and calligraphy fonts, which gives it the capability to reproduce unseen styles (both fonts and users' handwriting) in zero-shot. To the best of our knowledge, Emuru is the first autoregressive model for HTG, and the first designed specifically for generalization to novel styles. Moreover, our model generates images without background artifacts, which are easier to use for downstream applications. Extensive evaluation on both typewritten and handwritten, any-length text image generation scenarios demonstrates the effectiveness of our approach.
Authors: Victor Besnier, Mickael Chen, David Hurych, Eduardo Valle, Matthieu Cord
Abstract: Masked Generative Image Transformers (MaskGIT) have emerged as a scalable and efficient image generation framework, able to deliver high-quality visuals with low inference costs. However, MaskGIT's token unmasking scheduler, an essential component of the framework, has not received the attention it deserves. We analyze the sampling objective in MaskGIT, based on the mutual information between tokens, and elucidate its shortcomings. We then propose a new sampling strategy based on our Halton scheduler instead of the original Confidence scheduler. More precisely, our method selects the token's position according to a quasi-random, low-discrepancy Halton sequence. Intuitively, that method spreads the tokens spatially, progressively covering the image uniformly at each step. Our analysis shows that it allows reducing non-recoverable sampling errors, leading to simpler hyper-parameters tuning and better quality images. Our scheduler does not require retraining or noise injection and may serve as a simple drop-in replacement for the original sampling strategy. Evaluation of both class-to-image synthesis on ImageNet and text-to-image generation on the COCO dataset demonstrates that the Halton scheduler outperforms the Confidence scheduler quantitatively by reducing the FID and qualitatively by generating more diverse and more detailed images. Our code is at https://github.com/valeoai/Halton-MaskGIT.
Authors: Gensheng Pei, Tao Chen, Yujia Wang, Xinhao Cai, Xiangbo Shu, Tianfei Zhou, Yazhou Yao
Abstract: The CLIP model has demonstrated significant advancements in aligning visual and language modalities through large-scale pre-training on image-text pairs, enabling strong zero-shot classification and retrieval capabilities on various domains. However, CLIP's training remains computationally intensive, with high demands on both data processing and memory. To address these challenges, recent masking strategies have emerged, focusing on the selective removal of image patches to improve training efficiency. Although effective, these methods often compromise key semantic information, resulting in suboptimal alignment between visual features and text descriptions. In this work, we present a concise yet effective approach called Patch Generation-to-Selection to enhance CLIP's training efficiency while preserving critical semantic content. Our method introduces a gradual masking process in which a small set of candidate patches is first pre-selected as potential mask regions. Then, we apply Sobel edge detection across the entire image to generate an edge mask that prioritizes the retention of the primary object areas. Finally, similarity scores between the candidate mask patches and their neighboring patches are computed, with optimal transport normalization refining the selection process to ensure a balanced similarity matrix. Our approach, CLIP-PGS, sets new state-of-the-art results in zero-shot classification and retrieval tasks, achieving superior performance in robustness evaluation and language compositionality benchmarks.
Authors: Johan Edstedt, Andr\'e Mateus, Alberto Jaenal
Abstract: Structure-from-Motion (SfM) is the task of estimating 3D structure and camera poses from images. We define Collaborative SfM (ColabSfM) as sharing distributed SfM reconstructions. Sharing maps requires estimating a joint reference frame, which is typically referred to as registration. However, there is a lack of scalable methods and training datasets for registering SfM reconstructions. In this paper, we tackle this challenge by proposing the scalable task of point cloud registration for SfM reconstructions. We find that current registration methods cannot register SfM point clouds when trained on existing datasets. To this end, we propose a SfM registration dataset generation pipeline, leveraging partial reconstructions from synthetically generated camera trajectories for each scene. Finally, we propose a simple but impactful neural refiner on top of the SotA registration method RoITr that yields significant improvements, which we call RefineRoITr. Our extensive experimental evaluation shows that our proposed pipeline and model enables ColabSfM. Code is available at https://github.com/EricssonResearch/ColabSfM
Authors: Ruiyang Ha, Songyi Jiang, Bin Li, Bikang Pan, Yihang Zhu, Junjie Zhang, Xiatian Zhu, Shaogang Gong, Jingya Wang
Abstract: Conventional person re-identification (ReID) research is often limited to single-modality sensor data from static cameras, which fails to address the complexities of real-world scenarios where multi-modal signals are increasingly prevalent. For instance, consider an urban ReID system integrating stationary RGB cameras, nighttime infrared sensors, and UAVs equipped with dynamic tracking capabilities. Such systems face significant challenges due to variations in camera perspectives, lighting conditions, and sensor modalities, hindering effective person ReID. To address these challenges, we introduce the MP-ReID benchmark, a novel dataset designed specifically for multi-modality and multi-platform ReID. This benchmark uniquely compiles data from 1,930 identities across diverse modalities, including RGB, infrared, and thermal imaging, captured by both UAVs and ground-based cameras in indoor and outdoor environments. Building on this benchmark, we introduce Uni-Prompt ReID, a framework with specific-designed prompts, tailored for cross-modality and cross-platform scenarios. Our method consistently outperforms state-of-the-art approaches, establishing a robust foundation for future research in complex and dynamic ReID environments. Our dataset are available at:https://mp-reid.github.io/.
Authors: Boyuan Zheng, Shouyi Lu, Renbo Huang, Minqing Huang, Fan Lu, Wei Tian, Guirong Zhuo, Lu Xiong
Abstract: We introduce R2LDM, an innovative approach for generating dense and accurate 4D radar point clouds, guided by corresponding LiDAR point clouds. Instead of utilizing range images or bird's eye view (BEV) images, we represent both LiDAR and 4D radar point clouds using voxel features, which more effectively capture 3D shape information. Subsequently, we propose the Latent Voxel Diffusion Model (LVDM), which performs the diffusion process in the latent space. Additionally, a novel Latent Point Cloud Reconstruction (LPCR) module is utilized to reconstruct point clouds from high-dimensional latent voxel features. As a result, R2LDM effectively generates LiDAR-like point clouds from paired raw radar data. We evaluate our approach on two different datasets, and the experimental results demonstrate that our model achieves 6- to 10-fold densification of radar point clouds, outperforming state-of-the-art baselines in 4D radar point cloud super-resolution. Furthermore, the enhanced radar point clouds generated by our method significantly improve downstream tasks, achieving up to 31.7% improvement in point cloud registration recall rate and 24.9% improvement in object detection accuracy.
Authors: Yizhe Liu, Tong Jia, Da Cai, Hao Wang, Dongyue Chen
Abstract: Transparent and specular objects are frequently encountered in daily life, factories, and laboratories. However, due to the unique optical properties, the depth information on these objects is usually incomplete and inaccurate, which poses significant challenges for downstream robotics tasks. Therefore, it is crucial to accurately restore the depth information of transparent and specular objects. Previous depth completion methods for these objects usually use RGB information as an additional channel of the depth image to perform depth prediction. Due to the poor-texture characteristics of transparent and specular objects, these methods that rely heavily on color information tend to generate structure-less depth predictions. Moreover, these 2D methods cannot effectively explore the 3D structure hidden in the depth channel, resulting in depth ambiguity. To this end, we propose a geometry-aware assisted depth completion method for transparent and specular objects, which focuses on exploring the 3D structural cues of the scene. Specifically, besides extracting 2D features from RGB-D input, we back-project the input depth to a point cloud and build the 3D branch to extract hierarchical scene-level 3D structural features. To exploit 3D geometric information, we design several gated cross-modal fusion modules to effectively propagate multi-level 3D geometric features to the image branch. In addition, we propose an adaptive correlation aggregation strategy to appropriately assign 3D features to the corresponding 2D features. Extensive experiments on ClearGrasp, OOD, TransCG, and STD datasets show that our method outperforms other state-of-the-art methods. We further demonstrate that our method significantly enhances the performance of downstream robotic grasping tasks.
Authors: Yuanmin Tang, Jing Yu, Keke Gai, Jiamin Zhuang, Gang Xiong, Gaopeng Gou, Qi Wu
Abstract: Zero-Shot Composed Image Retrieval (ZS-CIR) involves diverse tasks with a broad range of visual content manipulation intent across domain, scene, object, and attribute. The key challenge for ZS-CIR tasks is to modify a reference image according to manipulation text to accurately retrieve a target image, especially when the reference image is missing essential target content. In this paper, we propose a novel prediction-based mapping network, named PrediCIR, to adaptively predict the missing target visual content in reference images in the latent space before mapping for accurate ZS-CIR. Specifically, a world view generation module first constructs a source view by omitting certain visual content of a target view, coupled with an action that includes the manipulation intent derived from existing image-caption pairs. Then, a target content prediction module trains a world model as a predictor to adaptively predict the missing visual information guided by user intention in manipulating text at the latent space. The two modules map an image with the predicted relevant information to a pseudo-word token without extra supervision. Our model shows strong generalization ability on six ZS-CIR tasks. It obtains consistent and significant performance boosts ranging from 1.73% to 4.45% over the best methods and achieves new state-of-the-art results on ZS-CIR. Our code is available at https://github.com/Pter61/predicir.
Authors: Robin Hesse, Do\u{g}ukan Ba\u{g}c{\i}, Bernt Schiele, Simone Schaub-Meyer, Stefan Roth
Abstract: Deep learning has become an essential part of computer vision, with deep neural networks (DNNs) excelling in predictive performance. However, they often fall short in other critical quality dimensions, such as robustness, calibration, or fairness. While existing studies have focused on a subset of these quality dimensions, none have explored a more general form of "well-behavedness" of DNNs. With this work, we address this gap by simultaneously studying nine different quality dimensions for image classification. Through a large-scale study, we provide a bird's-eye view by analyzing 326 backbone models and how different training paradigms and model architectures affect the quality dimensions. We reveal various new insights such that (i) vision-language models exhibit high fairness on ImageNet-1k classification and strong robustness against domain changes; (ii) self-supervised learning is an effective training paradigm to improve almost all considered quality dimensions; and (iii) the training dataset size is a major driver for most of the quality dimensions. We conclude our study by introducing the QUBA score (Quality Understanding Beyond Accuracy), a novel metric that ranks models across multiple dimensions of quality, enabling tailored recommendations based on specific user needs.
Authors: Jonas Mirlach, Lei Wan, Andreas Wiedholz, Hannan Ejaz Keen, Andreas Eich
Abstract: In autonomous driving, the integration of roadside perception systems is essential for overcoming occlusion challenges and enhancing the safety of Vulnerable Road Users (VRUs). While LiDAR and visual (RGB) sensors are commonly used, thermal imaging remains underrepresented in datasets, despite its acknowledged advantages for VRU detection in extreme lighting conditions. In this paper, we present R-LiViT, the first dataset to combine LiDAR, RGB, and thermal imaging from a roadside perspective, with a strong focus on VRUs. R-LiViT captures three intersections during both day and night, ensuring a diverse dataset. It includes 10,000 LiDAR frames and 2,400 temporally and spatially aligned RGB and thermal images across over 150 traffic scenarios, with 6 and 8 annotated classes respectively, providing a comprehensive resource for tasks such as object detection and tracking. The dataset1 and the code for reproducing our evaluation results2 are made publicly available.
Authors: Siyuan Yang, Shilin Lu, Shizheng Wang, Meng Hwa Er, Zengwei Zheng, Alex C. Kot
Abstract: This paper explores the promising interplay between spiking neural networks (SNNs) and event-based cameras for privacy-preserving human action recognition (HAR). The unique feature of event cameras in capturing only the outlines of motion, combined with SNNs' proficiency in processing spatiotemporal data through spikes, establishes a highly synergistic compatibility for event-based HAR. Previous studies, however, have been limited by SNNs' ability to process long-term temporal information, essential for precise HAR. In this paper, we introduce two novel frameworks to address this: temporal segment-based SNN (\textit{TS-SNN}) and 3D convolutional SNN (\textit{3D-SNN}). The \textit{TS-SNN} extracts long-term temporal information by dividing actions into shorter segments, while the \textit{3D-SNN} replaces 2D spatial elements with 3D components to facilitate the transmission of temporal information. To promote further research in event-based HAR, we create a dataset, \textit{FallingDetection-CeleX}, collected using the high-resolution CeleX-V event camera $(1280 \times 800)$, comprising 7 distinct actions. Extensive experimental results show that our proposed frameworks surpass state-of-the-art SNN methods on our newly collected dataset and three other neuromorphic datasets, showcasing their effectiveness in handling long-range temporal information for event-based HAR.
Authors: Davide Berasi, Matteo Farina, Massimiliano Mancini, Elisa Ricci, Nicola Strisciuglio
Abstract: Vision-Language Models (VLMs) learn a shared feature space for text and images, enabling the comparison of inputs of different modalities. While prior works demonstrated that VLMs organize natural language representations into regular structures encoding composite meanings, it remains unclear if compositional patterns also emerge in the visual embedding space. In this work, we investigate compositionality in the image domain, where the analysis of compositional properties is challenged by noise and sparsity of visual data. We address these problems and propose a framework, called Geodesically Decomposable Embeddings (GDE), that approximates image representations with geometry-aware compositional structures in the latent space. We demonstrate that visual embeddings of pre-trained VLMs exhibit a compositional arrangement, and evaluate the effectiveness of this property in the tasks of compositional classification and group robustness. GDE achieves stronger performance in compositional classification compared to its counterpart method that assumes linear geometry of the latent space. Notably, it is particularly effective for group robustness, where we achieve higher results than task-specific solutions. Our results indicate that VLMs can automatically develop a human-like form of compositional reasoning in the visual domain, making their underlying processes more interpretable. Code is available at https://github.com/BerasiDavide/vlm_image_compositionality.
URLs: https://github.com/BerasiDavide/vlm_image_compositionality.
Authors: Fouad Makiyeh, Huy-Dung Nguyen, Patrick Chareyre, Ramin Hasani, Marc Blanchon, Daniela Rus
Abstract: Steering estimation is a critical task in autonomous driving, traditionally relying on 2D image-based models. In this work, we explore the advantages of incorporating 3D spatial information through hybrid architectures that combine 3D neural network models with recurrent neural networks (RNNs) for temporal modeling, using LiDAR-based point clouds as input. We systematically evaluate four hybrid 3D models, all of which outperform the 2D-only baseline, with the Graph Neural Network (GNN) - RNN model yielding the best results. To reduce reliance on LiDAR, we leverage a pretrained unified model to estimate depth from monocular images, reconstructing pseudo-3D point clouds. We then adapt the GNN-RNN model, originally designed for LiDAR-based point clouds, to work with these pseudo-3D representations, achieving comparable or even superior performance compared to the LiDAR-based model. Additionally, the unified model provides semantic labels for each point, enabling a more structured scene representation. To further optimize graph construction, we introduce an efficient connectivity strategy where connections are predominantly formed between points of the same semantic class, with only 20\% of inter-class connections retained. This targeted approach reduces graph complexity and computational cost while preserving critical spatial relationships. Finally, we validate our approach on the KITTI dataset, achieving a 71% improvement over 2D-only models. Our findings highlight the advantages of 3D spatial information and efficient graph construction for steering estimation, while maintaining the cost-effectiveness of monocular images and avoiding the expense of LiDAR-based systems.
Authors: Panpan Wang, Liqiang Niu, Fandong Meng, Jinan Xu, Yufeng Chen, Jie Zhou
Abstract: In the domain of image generation, latent-based generative models occupy a dominant status; however, these models rely heavily on image tokenizer. To meet modeling requirements, autoregressive models possessing the characteristics of scalability and flexibility embrace a discrete-valued tokenizer, but face the challenge of poor image generation quality. In contrast, diffusion models take advantage of the continuous-valued tokenizer to achieve better generation quality but are subject to low efficiency and complexity. The existing hybrid models are mainly to compensate for information loss and simplify the diffusion learning process. The potential of merging discrete-valued and continuous-valued tokens in the field of image generation has not yet been explored. In this paper, we propose D2C, a novel two-stage method to enhance model generation capacity. In the first stage, the discrete-valued tokens representing coarse-grained image features are sampled by employing a small discrete-valued generator. Then in the second stage, the continuous-valued tokens representing fine-grained image features are learned conditioned on the discrete token sequence. In addition, we design two kinds of fusion modules for seamless interaction. On the ImageNet-256 benchmark, extensive experiment results validate that our model achieves superior performance compared with several continuous-valued and discrete-valued generative models on the class-conditional image generation tasks.
Authors: Tonmoy Hossain ana Miaomiao Zhang
Abstract: Deformable shape representations, parameterized by deformations relative to a given template, have proven effective for improved image analysis tasks. However, their broader applicability is hindered by two major challenges. First, existing methods mainly rely on a known template during testing, which is impractical and limits flexibility. Second, they often struggle to capture fine-grained, voxel-level distinctions between similar shapes (e.g., anatomical variations among healthy individuals, those with mild cognitive impairment, and diseased states). To address these limitations, we propose a novel framework - Contrastive Representation Learning of Deformable shapes (CoRLD) in learned deformation spaces and demonstrate its effectiveness in the context of image classification. Our CoRLD leverages a class-aware contrastive supervised learning objective in latent deformation spaces, promoting proximity among representations of similar classes while ensuring separation of dissimilar groups. In contrast to previous deep learning networks that require a reference image as input to predict deformation changes, our approach eliminates this dependency. Instead, template images are utilized solely as ground truth in the loss function during the training process, making our model more flexible and generalizable to a wide range of medical applications. We validate CoRLD on diverse datasets, including real brain magnetic resonance imaging (MRIs) and adrenal shapes derived from computed tomography (CT) scans. Experimental results show that our model effectively extracts deformable shape features, which can be easily integrated with existing classifiers to substantially boost the classification accuracy. Our code is available at GitHub.
Authors: Alexandra Arzberger, Ramin Tavakoli Kolagari
Abstract: Light Detection and Ranging (LiDAR) is an essential sensor technology for autonomous driving as it can capture high-resolution 3D data. As 3D object detection systems (OD) can interpret such point cloud data, they play a key role in the driving decisions of autonomous vehicles. Consequently, such 3D OD must be robust against all types of perturbations and must therefore be extensively tested. One approach is the use of adversarial examples, which are small, sometimes sophisticated perturbations in the input data that change, i.e., falsify, the prediction of the OD. These perturbations are carefully designed based on the weaknesses of the OD. The robustness of the OD cannot be quantified with adversarial examples in general, because if the OD is vulnerable to a given attack, it is unclear whether this is due to the robustness of the OD or whether the attack algorithm produces particularly strong adversarial examples. The contribution of this work is Hi-ALPS -- Hierarchical Adversarial-example-based LiDAR Perturbation Level System, where higher robustness of the OD is required to withstand the perturbations as the perturbation levels increase. In doing so, the Hi-ALPS levels successively implement a heuristic followed by established adversarial example approaches. In a series of comprehensive experiments using Hi-ALPS, we quantify the robustness of six state-of-the-art 3D OD under different types of perturbations. The results of the experiments show that none of the OD is robust against all Hi-ALPS levels; an important factor for the ranking is that human observers can still correctly recognize the perturbed objects, as the respective perturbations are small. To increase the robustness of the OD, we discuss the applicability of state-of-the-art countermeasures. In addition, we derive further suggestions for countermeasures based on our experimental results.
Authors: Duanrui Yu, Jing You, Xin Pei, Anqi Qu, Dingyu Wang, Shaocheng Jia
Abstract: Collaborative perception allows real-time inter-agent information exchange and thus offers invaluable opportunities to enhance the perception capabilities of individual agents. However, limited communication bandwidth in practical scenarios restricts the inter-agent data transmission volume, consequently resulting in performance declines in collaborative perception systems. This implies a trade-off between perception performance and communication cost. To address this issue, we propose Which2comm, a novel multi-agent 3D object detection framework leveraging object-level sparse features. By integrating semantic information of objects into 3D object detection boxes, we introduce semantic detection boxes (SemDBs). Innovatively transmitting these information-rich object-level sparse features among agents not only significantly reduces the demanding communication volume, but also improves 3D object detection performance. Specifically, a fully sparse network is constructed to extract SemDBs from individual agents; a temporal fusion approach with a relative temporal encoding mechanism is utilized to obtain the comprehensive spatiotemporal features. Extensive experiments on the V2XSet and OPV2V datasets demonstrate that Which2comm consistently outperforms other state-of-the-art methods on both perception performance and communication cost, exhibiting better robustness to real-world latency. These results present that for multi-agent collaborative 3D object detection, transmitting only object-level sparse features is sufficient to achieve high-precision and robust performance.
Authors: Patrick Rim, Hyoungseob Park, Vadim Ezhov, Jeffrey Moon, Alex Wong
Abstract: We propose PolyRad, a novel radar-guided depth estimation method that introduces polynomial fitting to transform scaleless depth predictions from pretrained monocular depth estimation (MDE) models into metric depth maps. Unlike existing approaches that rely on complex architectures or expensive sensors, our method is grounded in a simple yet fundamental insight: using polynomial coefficients predicted from cheap, ubiquitous radar data to adaptively adjust depth predictions non-uniformly across depth ranges. Although MDE models often infer reasonably accurate local depth structure within each object or local region, they may misalign these regions relative to one another, making a linear scale-and-shift transformation insufficient given three or more of these regions. In contrast, PolyRad generalizes beyond linear transformations and is able to correct such misalignments by introducing inflection points. Importantly, our polynomial fitting framework preserves structural consistency through a novel training objective that enforces monotonicity via first-derivative regularization. PolyRad achieves state-of-the-art performance on the nuScenes, ZJU-4DRadarCam, and View-of-Delft datasets, outperforming existing methods by 30.3% in MAE and 37.2% in RMSE.
Authors: Xueqi Qiu, Xingyu Miao, Fan Wan, Haoran Duan, Tejal Shah, Varun Ojhab, Yang Longa, Rajiv Ranjan
Abstract: Deepfake detection is crucial for curbing the harm it causes to society. However, current Deepfake detection methods fail to thoroughly explore artifact information across different domains due to insufficient intrinsic interactions. These interactions refer to the fusion and coordination after feature extraction processes across different domains, which are crucial for recognizing complex forgery clues. Focusing on more generalized Deepfake detection, in this work, we introduce a novel bi-directional attention module to capture the local positional information of artifact clues from the spatial domain. This enables accurate artifact localization, thus addressing the coarse processing with artifact features. To further address the limitation that the proposed bi-directional attention module may not well capture global subtle forgery information in the artifact feature (e.g., textures or edges), we employ a fine-grained frequency attention module in the frequency domain. By doing so, we can obtain high-frequency information in the fine-grained features, which contains the global and subtle forgery information. Although these features from the diverse domains can be effectively and independently improved, fusing them directly does not effectively improve the detection performance. Therefore, we propose a feature superposition strategy that complements information from spatial and frequency domains. This strategy turns the feature components into the form of wave-like tokens, which are updated based on their phase, such that the distinctions between authentic and artifact features can be amplified. Our method demonstrates significant improvements over state-of-the-art (SOTA) methods on five public Deepfake datasets in capturing abnormalities across different manipulated operations and real-life.
Authors: Xiaojin Lu, Taoran yue, Jiaxi cai, Shibing Chu
Abstract: Detecting infrared small targets in complex backgrounds remains a challenging task because of the low contrast and high noise levels inherent in infrared images. These factors often lead to the loss of crucial details during feature extraction. Moreover, existing detection methods have limitations in adequately integrating global and local information, which constrains the efficiency and accuracy of infrared small target detection. To address these challenges, this paper proposes a novel network architecture named MSCA-Net, which integrates three key components: Multi-Scale Enhanced Detection Attention mechanism(MSEDA), Positional Convolutional Block Attention Module (PCBAM), and Channel Aggregation Block (CAB). Specifically, MSEDA employs a multi-scale feature fusion attention mechanism to adaptively aggregate information across different scales, enriching feature representation. PCBAM captures the correlation between global and local features through a correlation matrix-based strategy, enabling deep feature interaction. Moreover, CAB redistributes input feature channels, facilitating the efficient transmission of beneficial features and further enhancing the model detection capability in complex backgrounds. The experimental results demonstrate that MSCA-Net achieves outstanding small target detection performance in complex backgrounds. Specifically, it attains mIoU scores of 78.43\%, 94.56\%, and 67.08\% on the NUAA-SIRST, NUDT-SIRST, and IRTSD-1K datasets, respectively, underscoring its effectiveness and strong potential for real-world applications.
Authors: Xingchao Yang, Takafumi Taketomi, Yuki Endo, Yoshihiro Kanamori
Abstract: Recovering high-quality 3D facial textures from single-view 2D images is a challenging task, especially under constraints of limited data and complex facial details such as makeup, wrinkles, and occlusions. In this paper, we introduce FreeUV, a novel ground-truth-free UV texture recovery framework that eliminates the need for annotated or synthetic UV data. FreeUV leverages pre-trained stable diffusion model alongside a Cross-Assembly inference strategy to fulfill this objective. In FreeUV, separate networks are trained independently to focus on realistic appearance and structural consistency, and these networks are combined during inference to generate coherent textures. Our approach accurately captures intricate facial features and demonstrates robust performance across diverse poses and occlusions. Extensive experiments validate FreeUV's effectiveness, with results surpassing state-of-the-art methods in both quantitative and qualitative metrics. Additionally, FreeUV enables new applications, including local editing, facial feature interpolation, and multi-view texture recovery. By reducing data requirements, FreeUV offers a scalable solution for generating high-fidelity 3D facial textures suitable for real-world scenarios.
Authors: Matthew Kenely, Dylan Seychell, Carl James Debono, Chris Porter
Abstract: News outlets' competition for attention in news interfaces has highlighted the need for demographically-aware saliency prediction models. Despite recent advancements in saliency detection applied to user interfaces (UI), existing datasets are limited in size and demographic representation. We present a deep learning framework that enhances the SaRa (Saliency Ranking) model with DeepGaze IIE, improving Salient Object Ranking (SOR) performance by 10.7%. Our framework optimizes three key components: saliency map generation, grid segment scoring, and map normalization. Through a two-fold experiment using eye-tracking (30 participants) and mouse-tracking (375 participants aged 13--70), we analyze attention patterns across demographic groups. Statistical analysis reveals significant age-based variations (p < 0.05, {\epsilon^2} = 0.042), with older users (36--70) engaging more with textual content and younger users (13--35) interacting more with images. Mouse-tracking data closely approximates eye-tracking behavior (sAUC = 0.86) and identifies UI elements that immediately stand out, validating its use in large-scale studies. We conclude that saliency studies should prioritize gathering data from a larger, demographically representative sample and report exact demographic distributions.
Authors: Ting Sun, Cheng Cui, Yuning Du, Yi Liu
Abstract: Document layout analysis is a critical preprocessing step in document intelligence, enabling the detection and localization of structural elements such as titles, text blocks, tables, and formulas. Despite its importance, existing layout detection models face significant challenges in generalizing across diverse document types, handling complex layouts, and achieving real-time performance for large-scale data processing. To address these limitations, we present PP-DocLayout, which achieves high precision and efficiency in recognizing 23 types of layout regions across diverse document formats. To meet different needs, we offer three models of varying scales. PP-DocLayout-L is a high-precision model based on the RT-DETR-L detector, achieving 90.4% mAP@0.5 and an end-to-end inference time of 13.4 ms per page on a T4 GPU. PP-DocLayout-M is a balanced model, offering 75.2% mAP@0.5 with an inference time of 12.7 ms per page on a T4 GPU. PP-DocLayout-S is a high-efficiency model designed for resource-constrained environments and real-time applications, with an inference time of 8.1 ms per page on a T4 GPU and 14.5 ms on a CPU. This work not only advances the state of the art in document layout analysis but also provides a robust solution for constructing high-quality training data, enabling advancements in document intelligence and multimodal AI systems. Code and models are available at https://github.com/PaddlePaddle/PaddleX .
Authors: Fanghua Yu, Jinjin Gu, Jinfan Hu, Zheyuan Li, Chao Dong
Abstract: We introduce UniCon, a novel architecture designed to enhance control and efficiency in training adapters for large-scale diffusion models. Unlike existing methods that rely on bidirectional interaction between the diffusion model and control adapter, UniCon implements a unidirectional flow from the diffusion network to the adapter, allowing the adapter alone to generate the final output. UniCon reduces computational demands by eliminating the need for the diffusion model to compute and store gradients during adapter training. Our results indicate that UniCon reduces GPU memory usage by one-third and increases training speed by 2.3 times, while maintaining the same adapter parameter size. Additionally, without requiring extra computational resources, UniCon enables the training of adapters with double the parameter volume of existing ControlNets. In a series of image conditional generation tasks, UniCon has demonstrated precise responsiveness to control inputs and exceptional generation capabilities.
Authors: Giacomo Savazzi, Eugenio Lomurno, Cristian Sbrolli, Agnese Chiatti, Matteo Matteucci
Abstract: As machine learning models increase in scale and complexity, obtaining sufficient training data has become a critical bottleneck due to acquisition costs, privacy constraints, and data scarcity in specialised domains. While synthetic data generation has emerged as a promising alternative, a notable performance gap remains compared to models trained on real data, particularly as task complexity grows. Concurrently, Neuro-Symbolic methods, which combine neural networks' learning strengths with symbolic reasoning's structured representations, have demonstrated significant potential across various cognitive tasks. This paper explores the utility of Neuro-Symbolic conditioning for synthetic image dataset generation, focusing specifically on improving the performance of Scene Graph Generation models. The research investigates whether structured symbolic representations in the form of scene graphs can enhance synthetic data quality through explicit encoding of relational constraints. The results demonstrate that Neuro-Symbolic conditioning yields significant improvements of up to +2.59% in standard Recall metrics and +2.83% in No Graph Constraint Recall metrics when used for dataset augmentation. These findings establish that merging Neuro-Symbolic and generative approaches produces synthetic data with complementary structural information that enhances model performance when combined with real data, providing a novel approach to overcome data scarcity limitations even for complex visual reasoning tasks.
Authors: Aryan Yazdan Parast, Basim Azam, Naveed Akhtar
Abstract: Deep neural networks trained with Empirical Risk Minimization (ERM) perform well when both training and test data come from the same domain, but they often fail to generalize to out-of-distribution samples. In image classification, these models may rely on spurious correlations that often exist between labels and irrelevant features of images, making predictions unreliable when those features do not exist. We propose a technique to generate training samples with text-to-image (T2I) diffusion models for addressing the spurious correlation problem. First, we compute the best describing token for the visual features pertaining to the causal components of samples by a textual inversion mechanism. Then, leveraging a language segmentation method and a diffusion model, we generate new samples by combining the causal component with the elements from other classes. We also meticulously prune the generated samples based on the prediction probabilities and attribution scores of the ERM model to ensure their correct composition for our objective. Finally, we retrain the ERM model on our augmented dataset. This process reduces the model's reliance on spurious correlations by learning from carefully crafted samples for in which this correlation does not exist. Our experiments show that across different benchmarks, our technique achieves better worst-group accuracy than the existing state-of-the-art methods.
Authors: Yu-Hsi Chen
Abstract: Detecting and tracking multiple unmanned aerial vehicles (UAVs) in thermal infrared video is inherently challenging due to low contrast, environmental noise, and small target sizes. This paper provides a straightforward approach to address multi-UAV tracking in thermal infrared video, leveraging recent advances in detection and tracking. Instead of relying on the YOLOv5 with the DeepSORT pipeline, we present a tracking framework built on YOLOv12 and BoT-SORT, enhanced with tailored training and inference strategies. We evaluate our approach following the metrics from the 4th Anti-UAV Challenge and demonstrate competitive performance. Notably, we achieve strong results without using contrast enhancement or temporal information fusion to enrich UAV features, highlighting our approach as a "Strong Baseline" for the multi-UAV tracking task. We provide implementation details, in-depth experimental analysis, and a discussion of potential improvements. The code is available at https://github.com/wish44165/YOLOv12-BoT-SORT-ReID .
Authors: Devavrat Tomar, Guillaume Vray, Dwarikanath Mahapatra, Sudipta Roy, Jean-Philippe Thiran, Behzad Bozorgtabar
Abstract: In this paper, we address the challenge of few-shot classification in histopathology whole slide images (WSIs) by utilizing foundational vision-language models (VLMs) and slide-level prompt learning. Given the gigapixel scale of WSIs, conventional multiple instance learning (MIL) methods rely on aggregation functions to derive slide-level (bag-level) predictions from patch representations, which require extensive bag-level labels for training. In contrast, VLM-based approaches excel at aligning visual embeddings of patches with candidate class text prompts but lack essential pathological prior knowledge. Our method distinguishes itself by utilizing pathological prior knowledge from language models to identify crucial local tissue types (patches) for WSI classification, integrating this within a VLM-based MIL framework. Our approach effectively aligns patch images with tissue types, and we fine-tune our model via prompt learning using only a few labeled WSIs per category. Experimentation on real-world pathological WSI datasets and ablation studies highlight our method's superior performance over existing MIL- and VLM-based methods in few-shot WSI classification tasks. Our code is publicly available at https://github.com/LTS5/SLIP.
Authors: Shuang Guo, Friedhelm Hamann, Guillermo Gallego
Abstract: Event cameras rely on motion to obtain information about scene appearance. In other words, for event cameras, motion and appearance are seen both or neither, which are encoded in the output event stream. Previous works consider recovering these two visual quantities as separate tasks, which does not fit with the nature of event cameras and neglects the inherent relations between both tasks. In this paper, we propose an unsupervised learning framework that jointly estimates optical flow (motion) and image intensity (appearance), with a single network. Starting from the event generation model, we newly derive the event-based photometric error as a function of optical flow and image intensity, which is further combined with the contrast maximization framework, yielding a comprehensive loss function that provides proper constraints for both flow and intensity estimation. Exhaustive experiments show that our model achieves state-of-the-art performance for both optical flow (achieves 20% and 25% improvement in EPE and AE respectively in the unsupervised learning category) and intensity estimation (produces competitive results with other baselines, particularly in high dynamic range scenarios). Last but not least, our model achieves shorter inference time than all the other optical flow models and many of the image reconstruction models, while they output only one quantity. Project page: https://github.com/tub-rip/e2fai
Authors: Hiromu Taketsugu, Takeru Oba, Takahiro Maeda, Shohei Nobuhara, Norimichi Ukita
Abstract: Humans can predict future human trajectories even from momentary observations by using human pose-related cues. However, previous Human Trajectory Prediction (HTP) methods leverage the pose cues implicitly, resulting in implausible predictions. To address this, we propose Locomotion Embodiment, a framework that explicitly evaluates the physical plausibility of the predicted trajectory by locomotion generation under the laws of physics. While the plausibility of locomotion is learned with an indifferentiable physics simulator, it is replaced by our differentiable Locomotion Value function to train an HTP network in a data-driven manner. In particular, our proposed Embodied Locomotion loss is beneficial for efficiently training a stochastic HTP network using multiple heads. Furthermore, the Locomotion Value filter is proposed to filter out implausible trajectories at inference. Experiments demonstrate that our method enhances even the state-of-the-art HTP methods across diverse datasets and problem settings. Our code is available at: https://github.com/ImIntheMiddle/EmLoco.
Authors: Vineet R Shenoy, Suhas Lohit, Hassan Mansour, Rama Chellappa, Tim K. Marks
Abstract: Camera-based monitoring of vital signs, also known as imaging photoplethysmography (iPPG), has seen applications in driver-monitoring, perfusion assessment in surgical settings, affective computing, and more. iPPG involves sensing the underlying cardiac pulse from video of the skin and estimating vital signs such as the heart rate or a full pulse waveform. Some previous iPPG methods impose model-based sparse priors on the pulse signals and use iterative optimization for pulse wave recovery, while others use end-to-end black-box deep learning methods. In contrast, we introduce methods that combine signal processing and deep learning methods in an inverse problem framework. Our methods estimate the underlying pulse signal and heart rate from facial video by learning deep-network-based denoising operators that leverage deep algorithm unfolding and deep equilibrium models. Experiments show that our methods can denoise an acquired signal from the face and infer the correct underlying pulse rate, achieving state-of-the-art heart rate estimation performance on well-known benchmarks, all with less than one-fifth the number of learnable parameters as the closest competing method.
Authors: Maria Pilligua, Danna Xue, Javier Vazquez-Corral
Abstract: Decomposing a video into a layer-based representation is crucial for easy video editing for the creative industries, as it enables independent editing of specific layers. Existing video-layer decomposition models rely on implicit neural representations (INRs) trained independently for each video, making the process time-consuming when applied to new videos. Noticing this limitation, we propose a meta-learning strategy to learn a generic video decomposition model to speed up the training on new videos. Our model is based on a hypernetwork architecture which, given a video-encoder embedding, generates the parameters for a compact INR-based neural video decomposition model. Our strategy mitigates the problem of single-video overfitting and, importantly, shortens the convergence of video decomposition on new, unseen videos. Our code is available at: https://hypernvd.github.io/
Authors: Louis Y. Kim, Michelle Karker, Victoria Valledor, Seiyoung C. Lee, Karl F. Brzoska, Margaret Duff, Anthony Palladino
Abstract: Recent advances in open-vocabulary object detection models will enable Automatic Target Recognition systems to be sustainable and repurposed by non-technical end-users for a variety of applications or missions. New, and potentially nuanced, classes can be defined with natural language text descriptions in the field, immediately before runtime, without needing to retrain the model. We present an approach for improving non-technical users' natural language text descriptions of their desired targets of interest, using a combination of analysis techniques on the text embeddings, and proper combinations of embeddings for contrastive examples. We quantify the improvement that our feedback mechanism provides by demonstrating performance with multiple publicly-available open-vocabulary object detection models.
Authors: Xianghan Meng, Zhiyuan Huang, Wei He, Xianbiao Qi, Rong Xiao, Chun-Guang Li
Abstract: Subspace clustering is a classical unsupervised learning task, built on a basic assumption that high-dimensional data can be approximated by a union of subspaces (UoS). Nevertheless, the real-world data are often deviating from the UoS assumption. To address this challenge, state-of-the-art deep subspace clustering algorithms attempt to jointly learn UoS representations and self-expressive coefficients. However, the general framework of the existing algorithms suffers from a catastrophic feature collapse and lacks a theoretical guarantee to learn desired UoS representation. In this paper, we present a Principled fRamewOrk for Deep Subspace Clustering (PRO-DSC), which is designed to learn structured representations and self-expressive coefficients in a unified manner. Specifically, in PRO-DSC, we incorporate an effective regularization on the learned representations into the self-expressive model, prove that the regularized self-expressive model is able to prevent feature space collapse, and demonstrate that the learned optimal representations under certain condition lie on a union of orthogonal subspaces. Moreover, we provide a scalable and efficient approach to implement our PRO-DSC and conduct extensive experiments to verify our theoretical findings and demonstrate the superior performance of our proposed deep subspace clustering approach. The code is available at https://github.com/mengxianghan123/PRO-DSC.
Authors: Wonbong Jang, Philippe Weinzaepfel, Vincent Leroy, Lourdes Agapito, Jerome Revaud
Abstract: We present Pow3r, a novel large 3D vision regression model that is highly versatile in the input modalities it accepts. Unlike previous feed-forward models that lack any mechanism to exploit known camera or scene priors at test time, Pow3r incorporates any combination of auxiliary information such as intrinsics, relative pose, dense or sparse depth, alongside input images, within a single network. Building upon the recent DUSt3R paradigm, a transformer-based architecture that leverages powerful pre-training, our lightweight and versatile conditioning acts as additional guidance for the network to predict more accurate estimates when auxiliary information is available. During training we feed the model with random subsets of modalities at each iteration, which enables the model to operate under different levels of known priors at test time. This in turn opens up new capabilities, such as performing inference in native image resolution, or point-cloud completion. Our experiments on 3D reconstruction, depth completion, multi-view depth prediction, multi-view stereo, and multi-view pose estimation tasks yield state-of-the-art results and confirm the effectiveness of Pow3r at exploiting all available information. The project webpage is https://europe.naverlabs.com/pow3r.
Authors: Jichen Hu, Chen Yang, Zanwei Zhou, Jiemin Fang, Xiaokang Yang, Qi Tian, Wei Shen
Abstract: Reflection removal of a single image remains a highly challenging task due to the complex entanglement between target scenes and unwanted reflections. Despite significant progress, existing methods are hindered by the scarcity of high-quality, diverse data and insufficient restoration priors, resulting in limited generalization across various real-world scenarios. In this paper, we propose Dereflection Any Image, a comprehensive solution with an efficient data preparation pipeline and a generalizable model for robust reflection removal. First, we introduce a dataset named Diverse Reflection Removal (DRR) created by randomly rotating reflective mediums in target scenes, enabling variation of reflection angles and intensities, and setting a new benchmark in scale, quality, and diversity. Second, we propose a diffusion-based framework with one-step diffusion for deterministic outputs and fast inference. To ensure stable learning, we design a three-stage progressive training strategy, including reflection-invariant finetuning to encourage consistent outputs across varying reflection patterns that characterize our dataset. Extensive experiments show that our method achieves SOTA performance on both common benchmarks and challenging in-the-wild images, showing superior generalization across diverse real-world scenes.
Authors: Jianing Qi, Jiawei Liu, Hao Tang, Zhigang Zhu
Abstract: Vision-Language Models (VLMs) excel at identifying and describing objects but struggle with spatial reasoning such as accurately understanding the relative positions of objects. Inspired by the dual-pathway (ventral-dorsal) model of human vision, we investigate why VLMs fail spatial tasks despite strong object recognition capabilities. Our interpretability-driven analysis reveals a critical underlying cause: vision embeddings in VLMs are treated primarily as semantic ``bag-of-tokens," overshadowing subtle yet crucial positional cues due to their disproportionately large embedding norms. We validate this insight through extensive diagnostic experiments, demonstrating minimal performance impact when token orders or fine-grained spatial details are removed. Guided by these findings, we propose simple, interpretable interventions, including normalizing vision embedding norms and extracting mid-layer spatially rich features, to restore spatial awareness. Empirical results on both our synthetic data and standard benchmarks demonstrate improved spatial reasoning capabilities, highlighting the value of interpretability-informed design choices. Our study not only uncovers fundamental limitations in current VLM architectures but also provides actionable insights for enhancing structured perception of visual scenes.
Authors: Qingyu Shi, Jianzong Wu, Jinbin Bai, Jiangning Zhang, Lu Qi, Xiangtai Li, Yunhai Tong
Abstract: The motion transfer task involves transferring motion from a source video to newly generated videos, requiring the model to decouple motion from appearance. Previous diffusion-based methods primarily rely on separate spatial and temporal attention mechanisms within 3D U-Net. In contrast, state-of-the-art video Diffusion Transformers (DiT) models use 3D full attention, which does not explicitly separate temporal and spatial information. Thus, the interaction between spatial and temporal dimensions makes decoupling motion and appearance more challenging for DiT models. In this paper, we propose DeT, a method that adapts DiT models to improve motion transfer ability. Our approach introduces a simple yet effective temporal kernel to smooth DiT features along the temporal dimension, facilitating the decoupling of foreground motion from background appearance. Meanwhile, the temporal kernel effectively captures temporal variations in DiT features, which are closely related to motion. Moreover, we introduce explicit supervision along dense trajectories in the latent feature space to further enhance motion consistency. Additionally, we present MTBench, a general and challenging benchmark for motion transfer. We also introduce a hybrid motion fidelity metric that considers both the global and local motion similarity. Therefore, our work provides a more comprehensive evaluation than previous works. Extensive experiments on MTBench demonstrate that DeT achieves the best trade-off between motion fidelity and edit fidelity.
Authors: Vineet R. Shenoy, Shaoju Wu, Armand Comas, Tim K. Marks, Suhas Lohit, Hassan Mansour
Abstract: Remote estimation of vital signs enables health monitoring for situations in which contact-based devices are either not available, too intrusive, or too expensive. In this paper, we present a modular, interpretable pipeline for pulse signal estimation from video of the face that achieves state-of-the-art results on publicly available datasets.Our imaging photoplethysmography (iPPG) system consists of three modules: face and landmark detection, time-series extraction, and pulse signal/pulse rate estimation. Unlike many deep learning methods that make use of a single black-box model that maps directly from input video to output signal or heart rate, our modular approach enables each of the three parts of the pipeline to be interpreted individually. The pulse signal estimation module, which we call TURNIP (Time-Series U-Net with Recurrence for Noise-Robust Imaging Photoplethysmography), allows the system to faithfully reconstruct the underlying pulse signal waveform and uses it to measure heart rate and pulse rate variability metrics, even in the presence of motion. When parts of the face are occluded due to extreme head poses, our system explicitly detects such "self-occluded" regions and maintains estimation robustness despite the missing information. Our algorithm provides reliable heart rate estimates without the need for specialized sensors or contact with the skin, outperforming previous iPPG methods on both color (RGB) and near-infrared (NIR) datasets.
Authors: Yihe Deng, Hritik Bansal, Fan Yin, Nanyun Peng, Wei Wang, Kai-Wei Chang
Abstract: Recent advancements demonstrated by DeepSeek-R1 have shown that complex reasoning abilities in large language models (LLMs), including sophisticated behaviors such as self-verification and self-correction, can be achieved by RL with verifiable rewards and significantly improves model performance on challenging tasks such as AIME. Motivated by these findings, our study investigates whether similar reasoning capabilities can be successfully integrated into large vision-language models (LVLMs) and assesses their impact on challenging multimodal reasoning tasks. We consider an approach that iteratively leverages supervised fine-tuning (SFT) on lightweight training data and Reinforcement Learning (RL) to further improve model generalization. Initially, reasoning capabilities were distilled from pure-text R1 models by generating reasoning steps using high-quality captions of the images sourced from diverse visual datasets. Subsequently, iterative RL training further enhance reasoning skills, with each iteration's RL-improved model generating refined SFT datasets for the next round. This iterative process yielded OpenVLThinker, a LVLM exhibiting consistently improved reasoning performance on challenging benchmarks such as MathVista, MathVerse, and MathVision, demonstrating the potential of our strategy for robust vision-language reasoning. The code, model and data are held at https://github.com/yihedeng9/OpenVLThinker.
Authors: Jerred Chen, Ronald Clark
Abstract: In many robotics and VR/AR applications, fast camera motions cause a high level of motion blur, causing existing camera pose estimation methods to fail. In this work, we propose a novel framework that leverages motion blur as a rich cue for motion estimation rather than treating it as an unwanted artifact. Our approach works by predicting a dense motion flow field and a monocular depth map directly from a single motion-blurred image. We then recover the instantaneous camera velocity by solving a linear least squares problem under the small motion assumption. In essence, our method produces an IMU-like measurement that robustly captures fast and aggressive camera movements. To train our model, we construct a large-scale dataset with realistic synthetic motion blur derived from ScanNet++v2 and further refine our model by training end-to-end on real data using our fully differentiable pipeline. Extensive evaluations on real-world benchmarks demonstrate that our method achieves state-of-the-art angular and translational velocity estimates, outperforming current methods like MASt3R and COLMAP.
Authors: Jiwen Yu, Yiran Qin, Haoxuan Che, Quande Liu, Xintao Wang, Pengfei Wan, Di Zhang, Xihui Liu
Abstract: Modern game development faces significant challenges in creativity and cost due to predetermined content in traditional game engines. Recent breakthroughs in video generation models, capable of synthesizing realistic and interactive virtual environments, present an opportunity to revolutionize game creation. In this position paper, we propose Interactive Generative Video (IGV) as the foundation for Generative Game Engines (GGE), enabling unlimited novel content generation in next-generation gaming. GGE leverages IGV's unique strengths in unlimited high-quality content synthesis, physics-aware world modeling, user-controlled interactivity, long-term memory capabilities, and causal reasoning. We present a comprehensive framework detailing GGE's core modules and a hierarchical maturity roadmap (L0-L4) to guide its evolution. Our work charts a new course for game development in the AI era, envisioning a future where AI-powered generative systems fundamentally reshape how games are created and experienced.
Authors: Mahmoud Hamash, Md Raqib Khan, Peter Tiernan
Abstract: STEAM education integrates Science, Technology, Engineering, Arts, and Mathematics to foster creativity and problem-solving. However, students with visual impairments (VI) encounter significant challenges in programming and robotics, particularly in tracking robot movements and developing spatial awareness. This paper presents a framework that leverages pre-constructed robots and algorithms, such as maze-solving techniques, within an accessible learning environment. The proposed system employs Contrastive Language-Image Pre-training (CLIP) to process global camera-captured maze layouts, converting visual data into textual descriptions that generate spatial audio prompts in an Audio Virtual Reality (AVR) system. Students issue verbal commands, which are refined through CLIP, while robot-mounted stereo cameras provide real-time data processed via Simultaneous Localization and Mapping (SLAM) for continuous feedback. By integrating these technologies, the framework empowers VI students to develop coding skills and engage in complex problem-solving tasks. Beyond maze-solving applications, this approach demonstrates the broader potential of computer vision in special education, contributing to improved accessibility and learning experiences in STEAM disciplines.
Authors: Kunal Chavan, Keertan Balaji, Spoorti Barigidad, Samba Raju Chiluveru
Abstract: With an increasing demand for assistive technologies that promote the independence and mobility of visually impaired people, this study suggests an innovative real-time system that gives audio descriptions of a user's surroundings to improve situational awareness. The system acquires live video input and processes it with a quantized and fine-tuned Florence-2 big model, adjusted to 4-bit accuracy for efficient operation on low-power edge devices such as the NVIDIA Jetson Orin Nano. By transforming the video signal into frames with a 5-frame latency, the model provides rapid and contextually pertinent descriptions of objects, pedestrians, and barriers, together with their estimated distances. The system employs Parler TTS Mini, a lightweight and adaptable Text-to-Speech (TTS) solution, for efficient audio feedback. It accommodates 34 distinct speaker types and enables customization of speech tone, pace, and style to suit user requirements. This study examines the quantization and fine-tuning techniques utilized to modify the Florence-2 model for this application, illustrating how the integration of a compact model architecture with a versatile TTS component improves real-time performance and user experience. The proposed system is assessed based on its accuracy, efficiency, and usefulness, providing a viable option to aid vision-impaired users in navigating their surroundings securely and successfully.
Authors: Grigorii Khvatskii, Yong Suk Lee, Corey Angst, Maria Gibbs, Robert Landers, Nitesh V. Chawla
Abstract: This paper examines the performance of Multimodal LLMs (MLLMs) in skilled production work, with a focus on welding. Using a novel data set of real-world and online weld images, annotated by a domain expert, we evaluate the performance of two state-of-the-art MLLMs in assessing weld acceptability across three contexts: RV \& Marine, Aeronautical, and Farming. While both models perform better on online images, likely due to prior exposure or memorization, they also perform relatively well on unseen, real-world weld images. Additionally, we introduce WeldPrompt, a prompting strategy that combines Chain-of-Thought generation with in-context learning to mitigate hallucinations and improve reasoning. WeldPrompt improves model recall in certain contexts but exhibits inconsistent performance across others. These results underscore the limitations and potentials of MLLMs in high-stakes technical domains and highlight the importance of fine-tuning, domain-specific data, and more sophisticated prompting strategies to improve model reliability. The study opens avenues for further research into multimodal learning in industry applications.
Authors: Hanae Elmekki, Saidul Islam, Ahmed Alagha, Hani Sami, Amanda Spilkin, Ehsan Zakeri, Antonela Mariel Zanuttini, Jamal Bentahar, Lyes Kadem, Wen-Fang Xie, Philippe Pibarot, Rabeb Mizouni, Hadi Otrok, Shakti Singh, Azzam Mourad
Abstract: Medical Ultrasound (US) imaging has seen increasing demands over the past years, becoming one of the most preferred imaging modalities in clinical practice due to its affordability, portability, and real-time capabilities. However, it faces several challenges that limit its applicability, such as operator dependency, variability in interpretation, and limited resolution, which are amplified by the low availability of trained experts. This calls for the need of autonomous systems that are capable of reducing the dependency on humans for increased efficiency and throughput. Reinforcement Learning (RL) comes as a rapidly advancing field under Artificial Intelligence (AI) that allows the development of autonomous and intelligent agents that are capable of executing complex tasks through rewarded interactions with their environments. Existing surveys on advancements in the US scanning domain predominantly focus on partially autonomous solutions leveraging AI for scanning guidance, organ identification, plane recognition, and diagnosis. However, none of these surveys explore the intersection between the stages of the US process and the recent advancements in RL solutions. To bridge this gap, this review proposes a comprehensive taxonomy that integrates the stages of the US process with the RL development pipeline. This taxonomy not only highlights recent RL advancements in the US domain but also identifies unresolved challenges crucial for achieving fully autonomous US systems. This work aims to offer a thorough review of current research efforts, highlighting the potential of RL in building autonomous US solutions while identifying limitations and opportunities for further advancements in this field.
Authors: Sabeen Ahmed, Nathan Parker, Margaret Park, Daniel Jeong, Lauren Peres, Evan W. Davis, Jennifer B. Permuth, Erin Siegel, Matthew B. Schabath, Yasin Yilmaz, Ghulam Rasool
Abstract: Cancer cachexia is a common metabolic disorder characterized by severe muscle atrophy which is associated with poor prognosis and quality of life. Monitoring skeletal muscle area (SMA) longitudinally through computed tomography (CT) scans, an imaging modality routinely acquired in cancer care, is an effective way to identify and track this condition. However, existing tools often lack full automation and exhibit inconsistent accuracy, limiting their potential for integration into clinical workflows. To address these challenges, we developed SMAART-AI (Skeletal Muscle Assessment-Automated and Reliable Tool-based on AI), an end-to-end automated pipeline powered by deep learning models (nnU-Net 2D) trained on mid-third lumbar level CT images with 5-fold cross-validation, ensuring generalizability and robustness. SMAART-AI incorporates an uncertainty-based mechanism to flag high-error SMA predictions for expert review, enhancing reliability. We combined the SMA, skeletal muscle index, BMI, and clinical data to train a multi-layer perceptron (MLP) model designed to predict cachexia at the time of cancer diagnosis. Tested on the gastroesophageal cancer dataset, SMAART-AI achieved a Dice score of 97.80% +/- 0.93%, with SMA estimated across all four datasets in this study at a median absolute error of 2.48% compared to manual annotations with SliceOmatic. Uncertainty metrics-variance, entropy, and coefficient of variation-strongly correlated with SMA prediction errors (0.83, 0.76, and 0.73 respectively). The MLP model predicts cachexia with 79% precision, providing clinicians with a reliable tool for early diagnosis and intervention. By combining automation, accuracy, and uncertainty awareness, SMAART-AI bridges the gap between research and clinical application, offering a transformative approach to managing cancer cachexia.
Authors: Hadi Amini, Md Jueal Mia, Yasaman Saadati, Ahmed Imteaj, Seyedsina Nabavirazavi, Urmish Thakker, Md Zarif Hossain, Awal Ahmed Fime, S. S. Iyengar
Abstract: Language models (LMs) are machine learning models designed to predict linguistic patterns by estimating the probability of word sequences based on large-scale datasets, such as text. LMs have a wide range of applications in natural language processing (NLP) tasks, including autocomplete and machine translation. Although larger datasets typically enhance LM performance, scalability remains a challenge due to constraints in computational power and resources. Distributed computing strategies offer essential solutions for improving scalability and managing the growing computational demand. Further, the use of sensitive datasets in training and deployment raises significant privacy concerns. Recent research has focused on developing decentralized techniques to enable distributed training and inference while utilizing diverse computational resources and enabling edge AI. This paper presents a survey on distributed solutions for various LMs, including large language models (LLMs), vision language models (VLMs), multimodal LLMs (MLLMs), and small language models (SLMs). While LLMs focus on processing and generating text, MLLMs are designed to handle multiple modalities of data (e.g., text, images, and audio) and to integrate them for broader applications. To this end, this paper reviews key advancements across the MLLM pipeline, including distributed training, inference, fine-tuning, and deployment, while also identifying the contributions, limitations, and future areas of improvement. Further, it categorizes the literature based on six primary focus areas of decentralization. Our analysis describes gaps in current methodologies for enabling distributed solutions for LMs and outline future research directions, emphasizing the need for novel solutions to enhance the robustness and applicability of distributed LMs.
Authors: Julian Ziegler, Patrick Frenzel, Mirco Fuchs
Abstract: This work concerns itself with the task of reconstructing all edges of an arbitrary 3D wire-frame model projected to an image plane. We explore a bottom-up part-wise procedure undertaken by an RL agent to segment and reconstruct these 2D multipart objects. The environment's state is represented as a four-colour image, where different colours correspond to background, a target edge, a reconstruction line, and the overlap of both. At each step, the agent can transform the reconstruction line within a four-dimensional action space or terminate the episode using a specific termination action. To investigate the impact of reward function formulations, we tested episodic and incremental rewards, as well as combined approaches. Empirical results demonstrated that the latter yielded the most effective training performance. To further enhance efficiency and stability, we introduce curriculum learning strategies. First, an action-based curriculum was implemented, where the agent was initially restricted to a reduced action space, being able to only perform three of the five possible actions, before progressing to the full action space. Second, we test a task-based curriculum, where the agent first solves a simplified version of the problem before being presented with the full, more complex task. This second approach produced promising results, as the agent not only successfully transitioned from learning the simplified task to mastering the full task, but in doing so gained significant performance. This study demonstrates the potential of an iterative RL wire-frame reconstruction in two dimensions. By combining optimized reward function formulations with curriculum learning strategies, we achieved significant improvements in training success. The proposed methodology provides an effective framework for solving similar tasks and represents a promising direction for future research in the field.
Authors: Dana Cohen-Bar, Daniel Cohen-Or, Gal Chechik, Yoni Kasten
Abstract: As 3D content creation continues to grow, transferring semantic textures between 3D meshes remains a significant challenge in computer graphics. While recent methods leverage text-to-image diffusion models for texturing, they often struggle to preserve the appearance of the source texture during texture transfer. We present \ourmethod, a novel approach that learns a volumetric texture field from a single textured mesh by mapping semantic features to surface colors. Using an efficient triplane-based architecture, our method enables semantic-aware texture transfer to a novel target mesh. Despite training on just one example, it generalizes effectively to diverse shapes within the same category. Extensive evaluation on our newly created benchmark dataset shows that \ourmethod{} achieves superior texture transfer quality and fast inference times compared to existing methods. Our approach advances single-example texture transfer, providing a practical solution for maintaining visual coherence across related 3D models in applications like game development and simulation.
Authors: Yinchi Zhou, Huidong Xie, Menghua Xia, Qiong Liu, Bo Zhou, Tianqi Chen, Jun Hou, Liang Guo, Xinyuan Zheng, Hanzhong Wang, Biao Li, Axel Rominger, Kuangyu Shi, Nicha C. Dvorneka, Chi Liu
Abstract: Low-count positron emission tomography (LCPET) imaging can reduce patients' exposure to radiation but often suffers from increased image noise and reduced lesion detectability, necessitating effective denoising techniques. Diffusion models have shown promise in LCPET denoising for recovering degraded image quality. However, training such models requires large and diverse datasets, which are challenging to obtain in the medical domain. To address data scarcity and privacy concerns, we combine diffusion models with federated learning -- a decentralized training approach where models are trained individually at different sites, and their parameters are aggregated on a central server over multiple iterations. The variation in scanner types and image noise levels within and across institutions poses additional challenges for federated learning in LCPET denoising. In this study, we propose a novel noise-embedded federated learning diffusion model (Fed-NDIF) to address these challenges, leveraging a multicenter dataset and varying count levels. Our approach incorporates liver normalized standard deviation (NSTD) noise embedding into a 2.5D diffusion model and utilizes the Federated Averaging (FedAvg) algorithm to aggregate locally trained models into a global model, which is subsequently fine-tuned on local datasets to optimize performance and obtain personalized models. Extensive validation on datasets from the University of Bern, Ruijin Hospital in Shanghai, and Yale-New Haven Hospital demonstrates the superior performance of our method in enhancing image quality and improving lesion quantification. The Fed-NDIF model shows significant improvements in PSNR, SSIM, and NMSE of the entire 3D volume, as well as enhanced lesion detectability and quantification, compared to local diffusion models and federated UNet-based models.
Authors: Mihaela-Larisa Clement, M\'onika Farsang, Felix Resch, Radu Grosu
Abstract: Autonomous agents that rely purely on perception to make real-time control decisions require efficient and robust architectures. In this work, we demonstrate that augmenting RGB input with depth information significantly enhances our agents' ability to predict steering commands compared to using RGB alone. We benchmark lightweight recurrent controllers that leverage the fused RGB-D features for sequential decision-making. To train our models, we collect high-quality data using a small-scale autonomous car controlled by an expert driver via a physical steering wheel, capturing varying levels of steering difficulty. Our models, trained under diverse configurations, were successfully deployed on real hardware. Specifically, our findings reveal that the early fusion of depth data results in a highly robust controller, which remains effective even with frame drops and increased noise levels, without compromising the network's focus on the task.
Authors: Chiara Schiavo, Elena Camuffo, Leonardo Badia, Simone Milani
Abstract: 3D Gaussian Splatting (3DGS) has significantly improved the efficiency and realism of three-dimensional scene visualization in several applications, ranging from robotics to eXtended Reality (XR). This work presents SAGE (Semantic-Driven Adaptive Gaussian Splatting in Extended Reality), a novel framework designed to enhance the user experience by dynamically adapting the Level of Detail (LOD) of different 3DGS objects identified via a semantic segmentation. Experimental results demonstrate how SAGE effectively reduces memory and computational overhead while keeping a desired target visual quality, thus providing a powerful optimization for interactive XR applications.
Authors: Yancheng Cai, Ali Bozorgian, Maliha Ashraf, Robert Wanat, Rafa{\l} K. Mantiuk
Abstract: The perception of flicker has been a prominent concern in illumination and electronic display fields for over a century. Traditional approaches often rely on Critical Flicker Frequency (CFF), primarily suited for high-contrast (full-on, full-off) flicker. To tackle varying contrast flicker, the International Committee for Display Metrology (ICDM) introduced a Temporal Contrast Sensitivity Function TCSF$_{IDMS}$ within the Information Display Measurements Standard (IDMS). Nevertheless, this standard overlooks crucial parameters: luminance, eccentricity, and area. Existing models incorporating these parameters are inadequate for flicker detection, especially at low spatial frequencies. To address these limitations, we extend the TCSF$_{IDMS}$ and combine it with a new spatial probability summation model to incorporate the effects of luminance, eccentricity, and area (elaTCSF). We train the elaTCSF on various flicker detection datasets and establish the first variable refresh rate flicker detection dataset for further verification. Additionally, we contribute to resolving a longstanding debate on whether the flicker is more visible in peripheral vision. We demonstrate how elaTCSF can be used to predict flicker due to low-persistence in VR headsets, identify flicker-free VRR operational ranges, and determine flicker sensitivity in lighting design.
Authors: Zichen Geng, Zeeshan Hayder, Wei Liu, Ajmal Saeed Mian
Abstract: Text-driven Human-Object Interaction (Text-to-HOI) generation is an emerging field with applications in animation, video games, virtual reality, and robotics. A key challenge in HOI generation is maintaining interaction consistency in long sequences. Existing Text-to-Motion-based approaches, such as discrete motion tokenization, cannot be directly applied to HOI generation due to limited data in this domain and the complexity of the modality. To address the problem of interaction consistency in long sequences, we propose an autoregressive diffusion model (ARDHOI) that predicts the next continuous token. Specifically, we introduce a Contrastive Variational Autoencoder (cVAE) to learn a physically plausible space of continuous HOI tokens, thereby ensuring that generated human-object motions are realistic and natural. For generating sequences autoregressively, we develop a Mamba-based context encoder to capture and maintain consistent sequential actions. Additionally, we implement an MLP-based denoiser to generate the subsequent token conditioned on the encoded context. Our model has been evaluated on the OMOMO and BEHAVE datasets, where it outperforms existing state-of-the-art methods in terms of both performance and inference speed. This makes ARDHOI a robust and efficient solution for text-driven HOI tasks
Authors: Shunki Tatsumi, Ryo Hayakawa, Youji Iiguni
Abstract: In this paper, we propose a depth-aided color image inpainting method in the quaternion domain, called depth-aided low-rank quaternion matrix completion (D-LRQMC). In conventional quaternion-based inpainting techniques, the color image is expressed as a quaternion matrix by using the three imaginary parts as the color channels, whereas the real part is set to zero and has no information. Our approach incorporates depth information as the real part of the quaternion representations, leveraging the correlation between color and depth to improve the result of inpainting. In the proposed method, we first restore the observed image with the conventional LRQMC and estimate the depth of the restored result. We then incorporate the estimated depth into the real part of the observed image and perform LRQMC again. Simulation results demonstrate that the proposed D-LRQMC can improve restoration accuracy and visual quality for various images compared to the conventional LRQMC. These results suggest the effectiveness of the depth information for color image processing in quaternion domain.
Authors: Basar Demir, Soumitri Chattopadhyay, Thomas Hastings Greer, Boqi Chen, Marc Niethammer
Abstract: Medical vision foundational models are used for a wide variety of tasks, including medical image segmentation and registration. This work evaluates the ability of these models to predict disease progression using a simple linear probe. We hypothesize that intermediate layer features of segmentation models capture structural information, while those of registration models encode knowledge of change over time. Beyond demonstrating that these features are useful for disease progression prediction, we also show that registration model features do not require spatially aligned input images. However, for segmentation models, spatial alignment is essential for optimal performance. Our findings highlight the importance of spatial alignment and the utility of foundation model features for image registration.
Authors: Hou In Derek Pun, Hou In Ivan Tam, Austin T. Wang, Xiaoliang Huo, Angel X. Chang, Manolis Savva
Abstract: Despite advances in indoor 3D scene layout generation, synthesizing scenes with dense object arrangements remains challenging. Existing methods primarily focus on large furniture while neglecting smaller objects, resulting in unrealistically empty scenes. Those that place small objects typically do not honor arrangement specifications, resulting in largely random placement not following the text description. We present HSM, a hierarchical framework for indoor scene generation with dense object arrangements across spatial scales. Indoor scenes are inherently hierarchical, with surfaces supporting objects at different scales, from large furniture on floors to smaller objects on tables and shelves. HSM embraces this hierarchy and exploits recurring cross-scale spatial patterns to generate complex and realistic indoor scenes in a unified manner. Our experiments show that HSM outperforms existing methods by generating scenes that are more realistic and better conform to user input across room types and spatial configurations.
Authors: Yiqiang Cai, Yizhou Tan, Peihong Zhang, Yuxuan Liu, Shengchen Li, Xi Shao, Mark D. Plumbley
Abstract: Acoustic scene recordings are often collected from a diverse range of cities. Most existing acoustic scene classification (ASC) approaches focus on identifying common acoustic scene patterns across cities to enhance generalization. In contrast, we hypothesize that city-specific environmental and cultural differences in acoustic features are beneficial for the ASC task. In this paper, we introduce City2Scene, a novel framework that leverages city features to improve ASC. City2Scene transfers the city-specific knowledge from city classification models to a scene classification model using knowledge distillation. We evaluated City2Scene on the DCASE Challenge Task 1 datasets, where each audio clip is annotated with both scene and city labels. Experimental results demonstrate that city features provide valuable information for classifying scenes. By distilling the city-specific knowledge, City2Scene effectively improves accuracy for various state-of-the-art ASC backbone models, including both CNNs and Transformers.
Authors: Mengsay Loem, Taiju Hosaka
Abstract: Visual question answering (VQA) has emerged as a flexible approach for extracting specific pieces of information from document images. However, existing work typically queries each field in isolation, overlooking potential dependencies across multiple items. This paper investigates the merits of extracting multiple fields jointly versus separately. Through experiments on multiple large vision language models and datasets, we show that jointly extracting fields often improves accuracy, especially when the fields share strong numeric or contextual dependencies. We further analyze how performance scales with the number of requested items and use a regression based metric to quantify inter field relationships. Our results suggest that multi field prompts can mitigate confusion arising from similar surface forms and related numeric values, providing practical methods for designing robust VQA systems in document information extraction tasks.
Authors: Xuan Wang, Siyuan Liang, Dongping Liao, Han Fang, Aishan Liu, Xiaochun Cao, Yu-liang Lu, Ee-Chien Chang, Xitong Gao
Abstract: Institutions with limited data and computing resources often outsource model training to third-party providers in a semi-honest setting, assuming adherence to prescribed training protocols with pre-defined learning paradigm (e.g., supervised or semi-supervised learning). However, this practice can introduce severe security risks, as adversaries may poison the training data to embed backdoors into the resulting model. Existing detection approaches predominantly rely on statistical analyses, which often fail to maintain universally accurate detection accuracy across different learning paradigms. To address this challenge, we propose a unified backdoor detection framework in the semi-honest setting that exploits cross-examination of model inconsistencies between two independent service providers. Specifically, we integrate central kernel alignment to enable robust feature similarity measurements across different model architectures and learning paradigms, thereby facilitating precise recovery and identification of backdoor triggers. We further introduce backdoor fine-tuning sensitivity analysis to distinguish backdoor triggers from adversarial perturbations, substantially reducing false positives. Extensive experiments demonstrate that our method achieves superior detection performance, improving accuracy by 5.4%, 1.6%, and 11.9% over SoTA baselines across supervised, semi-supervised, and autoregressive learning tasks, respectively. Notably, it is the first to effectively detect backdoors in multimodal large language models, further highlighting its broad applicability and advancing secure deep learning.
Authors: Ji-Hoon Kim, Jeongsoo Choi, Jaehun Kim, Chaeyoung Jung, Joon Son Chung
Abstract: The objective of this study is to generate high-quality speech from silent talking face videos, a task also known as video-to-speech synthesis. A significant challenge in video-to-speech synthesis lies in the substantial modality gap between silent video and multi-faceted speech. In this paper, we propose a novel video-to-speech system that effectively bridges this modality gap, significantly enhancing the quality of synthesized speech. This is achieved by learning of hierarchical representations from video to speech. Specifically, we gradually transform silent video into acoustic feature spaces through three sequential stages -- content, timbre, and prosody modeling. In each stage, we align visual factors -- lip movements, face identity, and facial expressions -- with corresponding acoustic counterparts to ensure the seamless transformation. Additionally, to generate realistic and coherent speech from the visual representations, we employ a flow matching model that estimates direct trajectories from a simple prior distribution to the target speech distribution. Extensive experiments demonstrate that our method achieves exceptional generation quality comparable to real utterances, outperforming existing methods by a significant margin.
Authors: Zhe Hu, Jing Li, Yu Yin
Abstract: Embodied decision-making is fundamental for AI agents operating in real-world environments. While Visual Language Models (VLMs) have advanced this capability, they still struggle with complex decisions, particularly in human-centered situations that require deep reasoning about human needs and values. In this study, we systematically evaluate open-sourced VLMs on multimodal human-centered decision-making tasks. We find that LLMs receiving only textual descriptions unexpectedly outperform their VLM counterparts of similar scale that process actual images, suggesting that visual alignment may hinder VLM abilities. To address this challenge, we propose a novel text-only training approach with synthesized textual data. This method strengthens VLMs' language components and transfers the learned abilities to multimodal inference, eliminating the need for expensive image-text paired data. Furthermore, we show that VLMs can achieve substantial performance gains through self-improvement, using training data generated by their LLM counterparts rather than relying on larger teacher models like GPT-4. Our findings establish a more efficient and scalable approach to enhancing VLMs' human-centered decision-making capabilities, opening new avenues for optimizing VLMs through self-improvement mechanisms.
Authors: Ying Ming (Department of Radiology Peking Union Medical College Hospital Chinese Academy of Medical Sciences and Peking Union Medical College), Shaoze Luo (Research and Development Center Canon Medical Systems), Longfei Zhao (Research and Development Center Canon Medical Systems), Qiqi Xu (Research and Development Center Canon Medical Systems), Wei Song (Department of Radiology Peking Union Medical College Hospital Chinese Academy of Medical Sciences and Peking Union Medical College)
Abstract: Accurate segmentation of pulmonary vessels plays a very critical role in diagnosing and assessing various lung diseases. In clinical practice, diagnosis is typically carried out using CTPA images. However, there is a lack of high-precision pulmonary vessel segmentation algorithms for CTPA, and pulmonary vessel segmentation for NCCT poses an even greater challenge. In this study, we propose a 3D image segmentation algorithm for automated pulmonary vessel segmentation from both contrast and non-contrast CT images. In the network, we designed a Vessel Lumen Structure Optimization Module (VLSOM), which extracts the centerline of vessels and adjusts the weights based on the positional information and adds a Cl-Dice-Loss to supervise the stability of the vessels structure. In addition, we designed a method for generating vessel GT from CTPA to NCCT for training models that support both CTPA and NCCT. In this work, we used 427 sets of high-precision annotated CT data from multiple vendors and countries. Finally, our experimental model achieved Cl-Recall, Cl-DICE and Recall values of 0.879, 0.909, 0.934 (CTPA) and 0.928, 0.936, 0.955 (NCCT) respectively. This shows that our model has achieved good performance in both accuracy and completeness of pulmonary vessel segmentation. In clinical visual evaluation, our model also had good segmentation performance on various disease types and can assist doctors in medical diagnosis, verifying the great potential of this method in clinical application.
Authors: Aoting Zhang, Dongbao Yang, Chang Liu, Xiaopeng Hong, Yu Zhou
Abstract: Existing class incremental learning is mainly designed for single-label classification task, which is ill-equipped for multi-label scenarios due to the inherent contradiction of learning objectives for samples with incomplete labels. We argue that the main challenge to overcome this contradiction in multi-label class-incremental learning (MLCIL) lies in the model's inability to clearly distinguish between known and unknown knowledge. This ambiguity hinders the model's ability to retain historical knowledge, master current classes, and prepare for future learning simultaneously. In this paper, we target at specifying what is known or not to accommodate Historical, Current, and Prospective knowledge for MLCIL and propose a novel framework termed as HCP. Specifically, (i) we clarify the known classes by dynamic feature purification and recall enhancement with distribution prior, enhancing the precision and retention of known information. (ii) We design prospective knowledge mining to probe the unknown, preparing the model for future learning. Extensive experiments validate that our method effectively alleviates catastrophic forgetting in MLCIL, surpassing the previous state-of-the-art by 3.3% on average accuracy for MS-COCO B0-C10 setting without replay buffers.
Authors: David Mildenberger, Paul Hager, Daniel Rueckert, Martin J Menten
Abstract: Supervised contrastive learning (SupCon) has proven to be a powerful alternative to the standard cross-entropy loss for classification of multi-class balanced datasets. However, it struggles to learn well-conditioned representations of datasets with long-tailed class distributions. This problem is potentially exacerbated for binary imbalanced distributions, which are commonly encountered during many real-world problems such as medical diagnosis. In experiments on seven binary datasets of natural and medical images, we show that the performance of SupCon decreases with increasing class imbalance. To substantiate these findings, we introduce two novel metrics that evaluate the quality of the learned representation space. By measuring the class distribution in local neighborhoods, we are able to uncover structural deficiencies of the representation space that classical metrics cannot detect. Informed by these insights, we propose two new supervised contrastive learning strategies tailored to binary imbalanced datasets that improve the structure of the representation space and increase downstream classification accuracy over standard SupCon by up to 35%. We make our code available.
Authors: Snigdha Paul, Sambit Mallick, Anindya Sen
Abstract: Computer vision has transformed medical diagnosis, treatment, and research through advanced image processing and machine learning techniques. Fracture classification, a critical area in healthcare, has greatly benefited from these advancements, yet accurate detection is challenged by complex patterns and image noise. Bit plane slicing enhances medical images by reducing noise interference and extracting informative features. This research explores partial denoising techniques to provide practical solutions for improved fracture analysis, ultimately enhancing patient care. The study explores deep learning model DenseNet and handcrafted feature extraction. Decision Tree and Random Forest, were employed to train and evaluate distinct image representations. These include the original image, the concatenation of the four bit planes from the LSB as well as MSB, the fully denoised image, and an image consisting of 6 bit planes from MSB and 2 denoised bit planes from LSB. The purpose of forming these diverse image representations is to analyze SNR as well as classification accuracy and identify the bit planes that contain the most informative features. Moreover, the study delves into the significance of partial denoising techniques in preserving crucial features, leading to improvements in classification results. Notably, this study shows that employing the Random Forest classifier, the partially denoised image representation exhibited a testing accuracy of 95.61% surpassing the performance of other image representations. The outcomes of this research provide valuable insights into the development of efficient preprocessing, feature extraction and classification approaches for fracture identification. By enhancing diagnostic accuracy, these advancements hold the potential to positively impact patient care and overall medical outcomes.
Authors: Dongsheng Yang, Qianying Liu, Wataru Sato, Takashi Minato, Chaoran Liu, Shin'ya Nishida
Abstract: Automatic robotic facial expression generation is crucial for human-robot interaction, as handcrafted methods based on fixed joint configurations often yield rigid and unnatural behaviors. Although recent automated techniques reduce the need for manual tuning, they tend to fall short by not adequately bridging the gap between human preferences and model predictions-resulting in a deficiency of nuanced and realistic expressions due to limited degrees of freedom and insufficient perceptual integration. In this work, we propose a novel learning-to-rank framework that leverages human feedback to address this discrepancy and enhanced the expressiveness of robotic faces. Specifically, we conduct pairwise comparison annotations to collect human preference data and develop the Human Affective Pairwise Impressions (HAPI) model, a Siamese RankNet-based approach that refines expression evaluation. Results obtained via Bayesian Optimization and online expression survey on a 35-DOF android platform demonstrate that our approach produces significantly more realistic and socially resonant expressions of Anger, Happiness, and Surprise than those generated by baseline and expert-designed methods. This confirms that our framework effectively bridges the gap between human preferences and model predictions while robustly aligning robotic expression generation with human affective responses.
Authors: Fangyijie Wang, Kathleen M. Curran, Gu\'enol\'e Silvestre
Abstract: Accurate segmentation of ultrasound (US) images of the cervical muscles is crucial for precision healthcare. The demand for automatic computer-assisted methods is high. However, the scarcity of labeled data hinders the development of these methods. Advanced semi-supervised learning approaches have displayed promise in overcoming this challenge by utilizing labeled and unlabeled data. This study introduces a novel semi-supervised learning (SSL) framework that integrates dual neural networks. This SSL framework utilizes both networks to generate pseudo-labels and cross-supervise each other at the pixel level. Additionally, a self-supervised contrastive learning strategy is introduced, which employs a pair of deep representations to enhance feature learning capabilities, particularly on unlabeled data. Our framework demonstrates competitive performance in cervical segmentation tasks. Our codes are publicly available on https://github.com/13204942/SSL\_Cervical\_Segmentation.
URLs: https://github.com/13204942/SSL\_Cervical\_Segmentation.
Authors: Yongkang Cheng, Shaoli Huang, Xuelin Chen, Jifeng Ning, Mingming Gong
Abstract: Diffusion models have demonstrated remarkable synthesis quality and diversity in generating co-speech gestures. However, the computationally intensive sampling steps associated with diffusion models hinder their practicality in real-world applications. Hence, we present DIDiffGes, for a Decoupled Semi-Implicit Diffusion model-based framework, that can synthesize high-quality, expressive gestures from speech using only a few sampling steps. Our approach leverages Generative Adversarial Networks (GANs) to enable large-step sampling for diffusion model. We decouple gesture data into body and hands distributions and further decompose them into marginal and conditional distributions. GANs model the marginal distribution implicitly, while L2 reconstruction loss learns the conditional distributions exciplictly. This strategy enhances GAN training stability and ensures expressiveness of generated full-body gestures. Our framework also learns to denoise root noise conditioned on local body representation, guaranteeing stability and realism. DIDiffGes can generate gestures from speech with just 10 sampling steps, without compromising quality and expressiveness, reducing the number of sampling steps by a factor of 100 compared to existing methods. Our user study reveals that our method outperforms state-of-the-art approaches in human likeness, appropriateness, and style correctness. Project is https://cyk990422.github.io/DIDiffGes.
Authors: Tiarna Lee, Esther Puyol-Ant\'on, Bram Ruijsink, Miaojing Shi, Andrew P. King
Abstract: Artificial intelligence (AI) is increasingly being used for medical imaging tasks. However, there can be biases in the resulting models, particularly when they were trained using imbalanced training datasets. One such example has been the strong race bias effect in cardiac magnetic resonance (CMR) image segmentation models. Although this phenomenon has been reported in a number of publications, little is known about the effectiveness of bias mitigation algorithms in this domain. We aim to investigate the impact of common bias mitigation methods to address bias between Black and White subjects in AI-based CMR segmentation models. Specifically, we use oversampling, importance reweighing and Group DRO as well as combinations of these techniques to mitigate the race bias. Furthermore, motivated by recent findings on the root causes of AI-based CMR segmentation bias, we evaluate the same methods using models trained and evaluated on cropped CMR images. We find that bias can be mitigated using oversampling, significantly improving performance for the underrepresented Black subjects whilst not significantly reducing the majority White subjects' performance. Group DRO also improves performance for Black subjects but not significantly, while reweighing decreases performance for Black subjects. Using a combination of oversampling and Group DRO also improves performance for Black subjects but not significantly. Using cropped images increases performance for both races and reduces the bias, whilst adding oversampling as a bias mitigation technique with cropped images reduces the bias further.
Authors: Kwan Yun, Chaelin Kim, Hangyeul Shin, Junyong Noh
Abstract: Recent 3D face editing methods using masks have produced high-quality edited images by leveraging Neural Radiance Fields (NeRF). Despite their impressive performance, existing methods often provide limited user control due to the use of pre-trained segmentation masks. To utilize masks with a desired layout, an extensive training dataset is required, which is challenging to gather. We present FFaceNeRF, a NeRF-based face editing technique that can overcome the challenge of limited user control due to the use of fixed mask layouts. Our method employs a geometry adapter with feature injection, allowing for effective manipulation of geometry attributes. Additionally, we adopt latent mixing for tri-plane augmentation, which enables training with a few samples. This facilitates rapid model adaptation to desired mask layouts, crucial for applications in fields like personalized medical imaging or creative face editing. Our comparative evaluations demonstrate that FFaceNeRF surpasses existing mask based face editing methods in terms of flexibility, control, and generated image quality, paving the way for future advancements in customized and high-fidelity 3D face editing. The code is available on the {\href{https://kwanyun.github.io/FFaceNeRF_page/}{project-page}}.
Authors: Marco Usai, Andrea Loddo, Alessandra Perniciano, Maurizio Atzori, Cecilia Di Ruberto
Abstract: Gastric cancer ranks as the fifth most common and fourth most lethal cancer globally, with a dismal 5-year survival rate of approximately 20%. Despite extensive research on its pathobiology, the prognostic predictability remains inadequate, compounded by pathologists' high workload and potential diagnostic errors. Thus, automated, accurate histopathological diagnosis tools are crucial. This study employs Machine Learning and Deep Learning techniques to classify histopathological images into healthy and cancerous categories. Using handcrafted and deep features with shallow learning classifiers on the GasHisSDB dataset, we offer a comparative analysis and insights into the most robust and high-performing combinations of features and classifiers for distinguishing between normal and abnormal histopathological images without fine-tuning strategies. With the RF classifier, our approach can reach F1 of 93.4%, demonstrating its validity.
Authors: Davide Antonio Mura, Michela Pinna, Lorenzo Putzu, Andrea Loddo, Alessandra Perniciano, Olga Mulas, Cecilia Di Ruberto
Abstract: The detection of blood disorders often hinges upon the quantification of specific blood cell types. Variations in cell counts may indicate the presence of pathological conditions. Thus, the significance of developing precise automatic systems for blood cell enumeration is underscored. The investigation focuses on a novel approach termed DE-ViT. This methodology is employed in a Few-Shot paradigm, wherein training relies on a limited number of images. Two distinct datasets are utilised for experimental purposes: the Raabin-WBC dataset for Leukocyte detection and a local dataset for Schistocyte identification. In addition to the DE-ViT model, two baseline models, Faster R-CNN 50 and Faster R-CNN X 101, are employed, with their outcomes being compared against those of the proposed model. While DE-ViT has demonstrated state-of-the-art performance on the COCO and LVIS datasets, both baseline models surpassed its performance on the Raabin-WBC dataset. Moreover, only Faster R-CNN X 101 yielded satisfactory results on the SC-IDB. The observed disparities in performance may possibly be attributed to domain shift phenomena.
Authors: Luca Rossetto, Werner Bailer, Duc-Tien Dang-Nguyen, Graham Healy, Bj\"orn {\TH}\'or J\'onsson, Onanong Kongmeesub, Hoang-Bao Le, Stevan Rudinac, Klaus Sch\"offmann, Florian Spiess, Allie Tran, Minh-Triet Tran, Quang-Linh Tran, Cathal Gurrin
Abstract: Egocentric video has seen increased interest in recent years, as it is used in a range of areas. However, most existing datasets are limited to a single perspective. In this paper, we present the CASTLE 2024 dataset, a multimodal collection containing ego- and exo-centric (i.e., first- and third-person perspective) video and audio from 15 time-aligned sources, as well as other sensor streams and auxiliary data. The dataset was recorded by volunteer participants over four days in a fixed location and includes the point of view of 10 participants, with an additional 5 fixed cameras providing an exocentric perspective. The entire dataset contains over 600 hours of UHD video recorded at 50 frames per second. In contrast to other datasets, CASTLE 2024 does not contain any partial censoring, such as blurred faces or distorted audio. The dataset is available via https://castle-dataset.github.io/.
Authors: Th\'eo Bodrito, Olivier Flasseur, Julien Mairal, Jean Ponce, Maud Langlois, Anne-Marie Lagrange
Abstract: The search for exoplanets is an active field in astronomy, with direct imaging as one of the most challenging methods due to faint exoplanet signals buried within stronger residual starlight. Successful detection requires advanced image processing to separate the exoplanet signal from this nuisance component. This paper presents a novel statistical model that captures nuisance fluctuations using a multi-scale approach, leveraging problem symmetries and a joint spectral channel representation grounded in physical principles. Our model integrates into an interpretable, end-to-end learnable framework for simultaneous exoplanet detection and flux estimation. The proposed algorithm is evaluated against the state of the art using datasets from the SPHERE instrument operating at the Very Large Telescope (VLT). It significantly improves the precision-recall trade-off, notably on challenging datasets that are otherwise unusable by astronomers. The proposed approach is computationally efficient, robust to varying data quality, and well suited for large-scale observational surveys.
Authors: Yongli Xiang, Ziming Hong, Lina Yao, Dadong Wang, Tongliang Liu
Abstract: Non-transferable learning (NTL) has been proposed to protect model intellectual property (IP) by creating a "non-transferable barrier" to restrict generalization from authorized to unauthorized domains. Recently, well-designed attack, which restores the unauthorized-domain performance by fine-tuning NTL models on few authorized samples, highlights the security risks of NTL-based applications. However, such attack requires modifying model weights, thus being invalid in the black-box scenario. This raises a critical question: can we trust the security of NTL models deployed as black-box systems? In this work, we reveal the first loophole of black-box NTL models by proposing a novel attack method (dubbed as JailNTL) to jailbreak the non-transferable barrier through test-time data disguising. The main idea of JailNTL is to disguise unauthorized data so it can be identified as authorized by the NTL model, thereby bypassing the non-transferable barrier without modifying the NTL model weights. Specifically, JailNTL encourages unauthorized-domain disguising in two levels, including: (i) data-intrinsic disguising (DID) for eliminating domain discrepancy and preserving class-related content at the input-level, and (ii) model-guided disguising (MGD) for mitigating output-level statistics difference of the NTL model. Empirically, when attacking state-of-the-art (SOTA) NTL models in the black-box scenario, JailNTL achieves an accuracy increase of up to 55.7% in the unauthorized domain by using only 1% authorized samples, largely exceeding existing SOTA white-box attacks.
Authors: Zilin Dai, Lehong Wang, Fangzhou Lin, Yidong Wang, Zhigang Li, Kazunori D Yamada, Ziming Zhang, Wang Lu
Abstract: Real-world machine learning applications often struggle with two major challenges: distribution shift and label noise. Models tend to overfit by focusing on redundant and uninformative features in the training data, which makes it hard for them to generalize to the target domain. Noisy data worsens this problem by causing further overfitting to the noise, meaning that existing methods often fail to tell the difference between true, invariant features and misleading, spurious ones. To tackle these issues, we introduce Anchor Alignment and Adaptive Weighting (A3W). This new algorithm uses sample reweighting guided by natural language processing (NLP) anchors to extract more representative features. In simple terms, A3W leverages semantic representations from natural language models as a source of domain-invariant prior knowledge. Additionally, it employs a weighted loss function that adjusts each sample's contribution based on its similarity to the corresponding NLP anchor. This adjustment makes the model more robust to noisy labels. Extensive experiments on standard benchmark datasets show that A3W consistently outperforms state-of-the-art domain generalization methods, offering significant improvements in both accuracy and robustness across different datasets and noise levels.
Authors: Jyothi Rikhab Chand, Mathews Jacob
Abstract: Current end-to-end (E2E) and plug-and-play (PnP) image reconstruction algorithms approximate the maximum a posteriori (MAP) estimate but cannot offer sampling from the posterior distribution, like diffusion models. By contrast, it is challenging for diffusion models to be trained in an E2E fashion. This paper introduces a Deep End-to-End Posterior ENergy (DEEPEN) framework, which enables MAP estimation as well as sampling. We learn the parameters of the posterior, which is the sum of the data consistency error and the negative log-prior distribution, using maximum likelihood optimization in an E2E fashion. The proposed approach does not require algorithm unrolling, and hence has a smaller computational and memory footprint than current E2E methods, while it does not require contraction constraints typically needed by current PnP methods. Our results demonstrate that DEEPEN offers improved performance than current E2E and PnP models in the MAP setting, while it also offers faster sampling compared to diffusion models. In addition, the learned energy-based model is observed to be more robust to changes in image acquisition settings.
Authors: Jie Mei, Chenyu Lin, Yu Qiu, Yaonan Wang, Hui Zhang, Ziyang Wang, Dong Dai
Abstract: Lung cancer is a leading cause of cancer-related deaths globally. PET-CT is crucial for imaging lung tumors, providing essential metabolic and anatomical information, while it faces challenges such as poor image quality, motion artifacts, and complex tumor morphology. Deep learning-based models are expected to address these problems, however, existing small-scale and private datasets limit significant performance improvements for these methods. Hence, we introduce a large-scale PET-CT lung tumor segmentation dataset, termed PCLT20K, which comprises 21,930 pairs of PET-CT images from 605 patients. Furthermore, we propose a cross-modal interactive perception network with Mamba (CIPA) for lung tumor segmentation in PET-CT images. Specifically, we design a channel-wise rectification module (CRM) that implements a channel state space block across multi-modal features to learn correlated representations and helps filter out modality-specific noise. A dynamic cross-modality interaction module (DCIM) is designed to effectively integrate position and context information, which employs PET images to learn regional position information and serves as a bridge to assist in modeling the relationships between local features of CT images. Extensive experiments on a comprehensive benchmark demonstrate the effectiveness of our CIPA compared to the current state-of-the-art segmentation methods. We hope our research can provide more exploration opportunities for medical image segmentation. The dataset and code are available at https://github.com/mj129/CIPA.
Authors: Muhammad Ahmed Mohsin, Muhammad Jazib, Zeeshan Alam, Muhmmad Farhan Khan, Muhammad Saad, Muhammad Ali Jamshed
Abstract: In the evolving landscape of 6G networks, semantic communications are poised to revolutionize data transmission by prioritizing the transmission of semantic meaning over raw data accuracy. This paper presents a Vision Transformer (ViT)-based semantic communication framework that has been deliberately designed to achieve high semantic similarity during image transmission while simultaneously minimizing the demand for bandwidth. By equipping ViT as the encoder-decoder framework, the proposed architecture can proficiently encode images into a high semantic content at the transmitter and precisely reconstruct the images, considering real-world fading and noise consideration at the receiver. Building on the attention mechanisms inherent to ViTs, our model outperforms Convolution Neural Network (CNNs) and Generative Adversarial Networks (GANs) tailored for generating such images. The architecture based on the proposed ViT network achieves the Peak Signal-to-noise Ratio (PSNR) of 38 dB, which is higher than other Deep Learning (DL) approaches in maintaining semantic similarity across different communication environments. These findings establish our ViT-based approach as a significant breakthrough in semantic communications.
Authors: Francisco Tellez, Enrique Torres-Giese
Abstract: In this paper, we introduce a shape descriptor that we call "interior function". This is a Topological Data Analysis (TDA) based descriptor that refines previous descriptors for image analysis. Using this concept, we define subcomplex lacunarity, a new index that quantifies geometric characteristics of necrosis in tumors such as conglomeration. Building on this framework, we propose a set of indices to analyze necrotic morphology and construct a diagram that captures the distinct structural and geometric properties of necrotic regions in tumors. We present an application of this framework in the study of MRIs of Glioblastomas (GB). Using cluster analysis, we identify four distinct subtypes of Glioblastomas that reflect geometric properties of necrotic regions.
Authors: Congyi Fan, Jian Guan, Xuanjia Zhao, Dongli Xu, Youtian Lin, Tong Ye, Pengming Feng, Haiwei Pan
Abstract: Automatically generating natural, diverse and rhythmic human dance movements driven by music is vital for virtual reality and film industries. However, generating dance that naturally follows music remains a challenge, as existing methods lack proper beat alignment and exhibit unnatural motion dynamics. In this paper, we propose Danceba, a novel framework that leverages gating mechanism to enhance rhythm-aware feature representation for music-driven dance generation, which achieves highly aligned dance poses with enhanced rhythmic sensitivity. Specifically, we introduce Phase-Based Rhythm Extraction (PRE) to precisely extract rhythmic information from musical phase data, capitalizing on the intrinsic periodicity and temporal structures of music. Additionally, we propose Temporal-Gated Causal Attention (TGCA) to focus on global rhythmic features, ensuring that dance movements closely follow the musical rhythm. We also introduce Parallel Mamba Motion Modeling (PMMM) architecture to separately model upper and lower body motions along with musical features, thereby improving the naturalness and diversity of generated dance movements. Extensive experiments confirm that Danceba outperforms state-of-the-art methods, achieving significantly better rhythmic alignment and motion diversity. Project page: https://danceba.github.io/ .
Authors: Yin Lin, Yehansen Chen, Baocai Yin, Jinshui Hu, Bing Yin, Cong Liu, Zengfu Wang
Abstract: Recently, visual-language learning (VLL) has shown great potential in enhancing visual-based person re-identification (ReID). Existing VLL-based ReID methods typically focus on image-text feature alignment at the whole-body level, while neglecting supervision on fine-grained part features, thus lacking constraints for local feature semantic consistency. To this end, we propose Part-Informed Visual-language Learning ($\pi$-VL) to enhance fine-grained visual features with part-informed language supervisions for ReID tasks. Specifically, $\pi$-VL introduces a human parsing-guided prompt tuning strategy and a hierarchical visual-language alignment paradigm to ensure within-part feature semantic consistency. The former combines both identity labels and human parsing maps to constitute pixel-level text prompts, and the latter fuses multi-scale visual features with a light-weight auxiliary head to perform fine-grained image-text alignment. As a plug-and-play and inference-free solution, our $\pi$-VL achieves performance comparable to or better than state-of-the-art methods on four commonly used ReID benchmarks. Notably, it reports 91.0% Rank-1 and 76.9% mAP on the challenging MSMT17 database, without bells and whistles.
Authors: Zeyu Yun, Juexiao Zhang, Yann LeCun, Yubei Chen
Abstract: Unsupervised representation learning has seen tremendous progress. However, it is constrained by its reliance on domain specific stationarity and topology, a limitation not found in biological intelligence systems. For instance, unlike computer vision, human vision can process visual signals sampled from highly irregular and non-stationary sensors. We introduce a novel framework that learns from high-dimensional data without prior knowledge of stationarity and topology. Our model, abbreviated as URLOST, combines a learnable self-organizing layer, spectral clustering, and a masked autoencoder (MAE). We evaluate its effectiveness on three diverse data modalities including simulated biological vision data, neural recordings from the primary visual cortex, and gene expressions. Compared to state-of-the-art unsupervised learning methods like SimCLR and MAE, our model excels at learning meaningful representations across diverse modalities without knowing their stationarity or topology. It also outperforms other methods that are not dependent on these factors, setting a new benchmark in the field. We position this work as a step toward unsupervised learning methods capable of generalizing across diverse high-dimensional data modalities.
Authors: Samet Hicsonmez, Nermin Samet, Fidan Samet, Oguz Bakir, Emre Akbas, Pinar Duygulu
Abstract: In this paper, we explore a new domain for video-to-video translation. Motivated by the availability of animation movies that are adopted from illustrated books for children, we aim to stylize these videos with the style of the original illustrations. Current state-of-the-art video-to-video translation models rely on having a video sequence or a single style image to stylize an input video. We introduce a new problem for video stylizing where an unordered set of images are used. This is a challenging task for two reasons: i) we do not have the advantage of temporal consistency as in video sequences; ii) it is more difficult to obtain consistent styles for video frames from a set of unordered images compared to using a single image. Most of the video-to-video translation methods are built on an image-to-image translation model, and integrate additional networks such as optical flow, or temporal predictors to capture temporal relations. These additional networks make the model training and inference complicated and slow down the process. To ensure temporal coherency in video-to-video style transfer, we propose a new generator network with feature warping layers which overcomes the limitations of the previous methods. We show the effectiveness of our method on three datasets both qualitatively and quantitatively. Code and pretrained models are available at https://github.com/giddyyupp/wait.
Authors: Thanh-Dat Truong, Utsav Prabhu, Bhiksha Raj, Jackson Cothren, Khoa Luu
Abstract: Continual Learning in semantic scene segmentation aims to continually learn new unseen classes in dynamic environments while maintaining previously learned knowledge. Prior studies focused on modeling the catastrophic forgetting and background shift challenges in continual learning. However, fairness, another major challenge that causes unfair predictions leading to low performance among major and minor classes, still needs to be well addressed. In addition, prior methods have yet to model the unknown classes well, thus resulting in producing non-discriminative features among unknown classes. This work presents a novel Fairness Learning via Contrastive Attention Approach to continual learning in semantic scene understanding. In particular, we first introduce a new Fairness Contrastive Clustering loss to address the problems of catastrophic forgetting and fairness. Then, we propose an attention-based visual grammar approach to effectively model the background shift problem and unknown classes, producing better feature representations for different unknown classes. Through our experiments, our proposed approach achieves State-of-the-Art (SoTA) performance on different continual learning benchmarks, i.e., ADE20K, Cityscapes, and Pascal VOC. It promotes the fairness of the continual semantic segmentation model.
Authors: Kevin Klein, Antoine Muller, Alyssa Wohde, Alexander V. Gorelik, Volker Heyd, Ralf L\"ammel, Yoan Diekmann, Maxime Brami
Abstract: The context of this paper is the creation of large uniform archaeological datasets from heterogeneous published resources, such as find catalogues - with the help of AI and Big Data. The paper is concerned with the challenge of consistent assemblages of archaeological data. We cannot simply combine existing records, as they differ in terms of quality and recording standards. Thus, records have to be recreated from published archaeological illustrations. This is only a viable path with the help of automation. The contribution of this paper is a new workflow for collecting data from archaeological find catalogues available as legacy resources, such as archaeological drawings and photographs in large unsorted PDF files; the workflow relies on custom software (AutArch) supporting image processing, object detection, and interactive means of validating and adjusting automatically retrieved data. We integrate artificial intelligence (AI) in terms of neural networks for object detection and classification into the workflow, thereby speeding up, automating, and standardising data collection. Objects commonly found in archaeological catalogues - such as graves, skeletons, ceramics, ornaments, stone tools and maps - are detected. Those objects are spatially related and analysed to extract real-life attributes, such as the size and orientation of graves based on the north arrow and the scale. We also automate recording of geometric whole-outlines through contour detection, as an alternative to landmark-based geometric morphometrics. Detected objects, contours, and other automatically retrieved data can be manually validated and adjusted. We use third millennium BC Europe (encompassing cultures such as 'Corded Ware' and 'Bell Beaker', and their burial practices) as a 'testing ground' and for evaluation purposes; this includes a user study for the workflow and the AutArch software.
Authors: Taekyung Kim, Byeongho Heo, Dongyoon Han
Abstract: Masked image modeling (MIM) has emerged as a promising approach for pre-training Vision Transformers (ViTs). MIMs predict masked tokens token-wise to recover target signals that are tokenized from images or generated by pre-trained models like vision-language models. While using tokenizers or pre-trained models is viable, they often offer spatially inconsistent supervision even for neighboring tokens, hindering models from learning discriminative representations. Our pilot study identifies spatial inconsistency in supervisory signals and suggests that addressing it can improve representation learning. Building upon this insight, we introduce Dynamic Token Morphing (DTM), a novel method that dynamically aggregates tokens while preserving context to generate contextualized targets, thereby likely reducing spatial inconsistency. DTM is compatible with various SSL frameworks; we showcase significantly improved MIM results, barely introducing extra training costs. Our method facilitates MIM training by using more spatially consistent targets, resulting in improved training trends as evidenced by lower losses. Experiments on ImageNet-1K and ADE20K demonstrate DTM's superiority, which surpasses complex state-of-the-art MIM methods. Furthermore, the evaluation of transfer learning on downstream tasks like iNaturalist, along with extensive empirical studies, supports DTM's effectiveness.
Authors: Dongyang Liu, Renrui Zhang, Longtian Qiu, Siyuan Huang, Weifeng Lin, Shitian Zhao, Shijie Geng, Ziyi Lin, Peng Jin, Kaipeng Zhang, Wenqi Shao, Chao Xu, Conghui He, Junjun He, Hao Shao, Pan Lu, Hongsheng Li, Yu Qiao, Peng Gao
Abstract: We propose SPHINX-X, an extensive Multimodality Large Language Model (MLLM) series developed upon SPHINX. To improve the architecture and training efficiency, we modify the SPHINX framework by removing redundant visual encoders, bypassing fully-padded sub-images with skip tokens, and simplifying multi-stage training into a one-stage all-in-one paradigm. To fully unleash the potential of MLLMs, we assemble a comprehensive multi-domain and multimodal dataset covering publicly available resources in language, vision, and vision-language tasks. We further enrich this collection with our curated OCR intensive and Set-of-Mark datasets, extending the diversity and generality. By training over different base LLMs including TinyLlama1.1B, InternLM2-7B, LLaMA2-13B, and Mixtral8x7B, we obtain a spectrum of MLLMs that vary in parameter size and multilingual capabilities. Comprehensive benchmarking reveals a strong correlation between the multi-modal performance with the data and parameter scales. Code and models are released at https://github.com/Alpha-VLLM/LLaMA2-Accessory
Authors: Hamed Haghighi, Xiaomeng Wang, Hao Jing, Mehrdad Dianati
Abstract: Perception sensors, particularly camera and Lidar, are key elements of Autonomous Driving Systems (ADS) that enable them to comprehend their surroundings to informed driving and control decisions. Therefore, developing realistic simulation models for these sensors is essential for conducting effective simulation-based testing of ADS. Moreover, the rise of deep learning-based perception models has increased the utility of sensor simulation models for synthesising diverse training datasets. The traditional sensor simulation models rely on computationally expensive physics-based algorithms, specifically in complex systems such as ADS. Hence, the current potential resides in data-driven approaches, fuelled by the exceptional performance of deep generative models in capturing high-dimensional data distribution and volume renderers in accurately representing scenes. This paper reviews the current state-of-the-art data-driven camera and Lidar simulation models and their evaluation methods. It explores a spectrum of models from the novel perspective of generative models and volume renderers. Generative models are discussed in terms of their input-output types, while volume renderers are categorised based on their input encoding. Finally, the paper illustrates commonly used evaluation techniques for assessing sensor simulation models and highlights the existing research gaps in the area.
Authors: Enguang Wang, Zhimao Peng, Zhengyuan Xie, Fei Yang, Xialei Liu, Ming-Ming Cheng
Abstract: Given unlabelled datasets containing both old and new categories, generalized category discovery (GCD) aims to accurately discover new classes while correctly classifying old classes. Current GCD methods only use a single visual modality of information, resulting in a poor classification of visually similar classes. As a different modality, text information can provide complementary discriminative information, which motivates us to introduce it into the GCD task. However, the lack of class names for unlabelled data makes it impractical to utilize text information. To tackle this challenging problem, in this paper, we propose a Text Embedding Synthesizer (TES) to generate pseudo text embeddings for unlabelled samples. Specifically, our TES leverages the property that CLIP can generate aligned vision-language features, converting visual embeddings into tokens of the CLIP's text encoder to generate pseudo text embeddings. Besides, we employ a dual-branch framework, through the joint learning and instance consistency of different modality branches, visual and semantic information mutually enhance each other, promoting the interaction and fusion of visual and text knowledge. Our method unlocks the multi-modal potentials of CLIP and outperforms the baseline methods by a large margin on all GCD benchmarks, achieving new state-of-the-art. Our code is available at: https://github.com/enguangW/GET.
Authors: Jianing Yang, Xuweiyi Chen, Nikhil Madaan, Madhavan Iyengar, Shengyi Qian, David F. Fouhey, Joyce Chai
Abstract: The integration of language and 3D perception is crucial for embodied agents and robots that comprehend and interact with the physical world. While large language models (LLMs) have demonstrated impressive language understanding and generation capabilities, their adaptation to 3D environments (3D-LLMs) remains in its early stages. A primary challenge is a lack of large-scale datasets with dense grounding between language and 3D scenes. We introduce 3D-GRAND, a pioneering large-scale dataset comprising 40,087 household scenes paired with 6.2 million densely-grounded scene-language instructions. Our results show that instruction tuning with 3D-GRAND significantly enhances grounding capabilities and reduces hallucinations in 3D-LLMs. As part of our contributions, we propose a comprehensive benchmark 3D-POPE to systematically evaluate hallucination in 3D-LLMs, enabling fair comparisons of models. Our experiments highlight a scaling effect between dataset size and 3D-LLM performance, emphasizing the importance of large-scale 3D-text datasets for embodied AI research. Our results demonstrate early signals for effective sim-to-real transfer, indicating that models trained on large synthetic data can perform well on real-world 3D scans. Through 3D-GRAND and 3D-POPE, we aim to equip the embodied AI community with resources and insights to lead to more reliable and better-grounded 3D-LLMs. Project website: https://3d-grand.github.io
Authors: Jongwoo Park, Kanchana Ranasinghe, Kumara Kahatapitiya, Wonjeong Ryu, Donghyun Kim, Michael S. Ryoo
Abstract: Long-form videos that span across wide temporal intervals are highly information redundant and contain multiple distinct events or entities that are often loosely related. Therefore, when performing long-form video question answering (LVQA), all information necessary to generate a correct response can often be contained within a small subset of frames. Recent literature explore use of large language models (LLMs) in LVQA benchmarks, achieving exceptional performance, while relying on vision language models (VLMs) to convert all visual content within videos into natural language. Such VLMs often independently caption a large number of frames uniformly sampled from long videos, which is not efficient and can mostly be redundant. Questioning these decision choices, we explore optimal strategies for key-frame selection that can significantly reduce these redundancies, namely Hierarchical Keyframe Selector. Our proposed framework, LVNet, achieves state-of-the-art performance at a comparable caption scale across three benchmark LVQA datasets: EgoSchema, NExT-QA, and IntentQA, while also demonstrating a strong performance on videos up to an hour long in VideoMME. Our code will be released publicly. The code can be found at https://github.com/jongwoopark7978/LVNet.
Authors: Runze Liu, Dongchen Zhu, Guanghui Zhang, Lei Wang, Jiamao Li
Abstract: Self-supervised monocular depth estimation has received widespread attention because of its capability to train without ground truth. In real-world scenarios, the images may be blurry or noisy due to the influence of weather conditions and inherent limitations of the camera. Therefore, it is particularly important to develop a robust depth estimation model. Benefiting from the training strategies of generative networks, generative-based methods often exhibit enhanced robustness. In light of this, we employ the generative-based diffusion model with a unique denoising training process for self-supervised monocular depth estimation. Additionally, to further enhance the robustness of the diffusion model, we probe into the influence of perturbations on image features and propose a hierarchical feature-guided denoising module. Furthermore, we explore the implicit depth within reprojection and design an implicit depth consistency loss. This loss function is not interfered by the other subnetwork, which can be targeted to constrain the depth estimation network and ensure the scale consistency of depth within a video sequence. We conduct experiments on the KITTI and Make3D datasets. The results indicate that our approach stands out among generative-based models, while also showcasing remarkable robustness.
Authors: Anna Susmelj, Mael Macuglia, Nata\v{s}a Tagasovska, Reto Sutter, Sebastiano Caprara, Jean-Philippe Thiran, Ender Konukoglu
Abstract: Implicit functions such as Neural Radiance Fields (NeRFs), occupancy networks, and signed distance functions (SDFs) have become pivotal in computer vision for reconstructing detailed object shapes from sparse views. Achieving optimal performance with these models can be challenging due to the extreme sparsity of inputs and distribution shifts induced by data corruptions. To this end, large, noise-free synthetic datasets can serve as shape priors to help models fill in gaps, but the resulting reconstructions must be approached with caution. Uncertainty estimation is crucial for assessing the quality of these reconstructions, particularly in identifying areas where the model is uncertain about the parts it has inferred from the prior. In this paper, we introduce Dropsembles, a novel method for uncertainty estimation in tuned implicit functions. We demonstrate the efficacy of our approach through a series of experiments, starting with toy examples and progressing to a real-world scenario. Specifically, we train a Convolutional Occupancy Network on synthetic anatomical data and test it on low-resolution MRI segmentations of the lumbar spine. Our results show that Dropsembles achieve the accuracy and calibration levels of deep ensembles but with significantly less computational cost.
Authors: Filipe Gama, Matej Misar, Lukas Navara, Sergiu T. Popescu, Matej Hoffmann
Abstract: Automatic markerless estimation of infant posture and motion from ordinary videos carries great potential for movement studies "in the wild", facilitating understanding of motor development and massively increasing the chances of early diagnosis of disorders. There is rapid development of human pose estimation methods in computer vision thanks to advances in deep learning and machine learning. However, these methods are trained on datasets that feature adults in different contexts. This work tests and compares seven popular methods (AlphaPose, DeepLabCut/DeeperCut, Detectron2, HRNet, MediaPipe/BlazePose, OpenPose, and ViTPose) on videos of infants in supine position and in more complex settings. Surprisingly, all methods except DeepLabCut and MediaPipe have competitive performance without additional finetuning, with ViTPose performing best. Next to standard performance metrics (average precision and recall), we introduce errors expressed in the neck-mid-hip (torso length) ratio and additionally study missed and redundant detections, and the reliability of the internal confidence ratings of the different methods, which are relevant for downstream tasks. Among the networks with competitive performance, only AlphaPose could run close to real time (27 fps) on our machine. We provide documented Docker containers or instructions for all the methods we used, our analysis scripts, and the processed data at https://hub.docker.com/u/humanoidsctu and https://osf.io/x465b/.
URLs: https://hub.docker.com/u/humanoidsctu, https://osf.io/x465b/.
Authors: Wenqiang Zu, Shenghao Xie, Qing Zhao, Guoqi Li, Lei Ma
Abstract: Foundation models pre-trained on large-scale data have been widely witnessed to achieve success in various natural imaging downstream tasks. Parameter-efficient fine-tuning (PEFT) methods aim to adapt foundation models to new domains by updating only a small portion of parameters in order to reduce computational overhead. However, the effectiveness of these PEFT methods, especially in cross-domain few-shot scenarios, e.g., medical image analysis, has not been fully explored. In this work, we facilitate the study of the performance of PEFT when adapting foundation models to medical image classification tasks. Furthermore, to alleviate the limitations of prompt introducing ways and approximation capabilities on Transformer architectures of mainstream prompt tuning methods, we propose the Embedded Prompt Tuning (EPT) method by embedding prompt tokens into the expanded channels. We also find that there are anomalies in the feature space distribution of foundation models during pre-training process, and prompt tuning can help mitigate this negative impact. To explain this phenomenon, we also introduce a novel perspective to understand prompt tuning: Prompt tuning is a distribution calibrator. And we support it by analyzing patch-wise scaling and feature separation operations contained in EPT. Our experiments show that EPT outperforms several state-of-the-art fine-tuning methods by a significant margin on few-shot medical image classification tasks, and completes the fine-tuning process within highly competitive time, indicating EPT is an effective PEFT method. The source code is available at github.com/zuwenqiang/EPT.
Authors: Pascal J. Sager, Jan M. Deriu, Benjamin F. Grewe, Thilo Stadelmann, Christoph von der Malsburg
Abstract: We introduce the Cooperative Network Architecture (CNA), a model that represents sensory signals using structured, recurrently connected networks of neurons, termed "nets." Nets are dynamically assembled from overlapping net fragments, which are learned based on statistical regularities in sensory input. This architecture offers robustness to noise, deformation, and out-of-distribution data, addressing challenges in current vision systems from a novel perspective. We demonstrate that net fragments can be learned without supervision and flexibly recombined to encode novel patterns, enabling figure completion and resilience to noise. Our findings establish CNA as a promising paradigm for developing neural representations that integrate local feature processing with global structure formation, providing a foundation for future research on invariant object recognition.
Authors: Chinedu Innocent Nwoye, Rupak Bose, Kareem Elgohary, Lorenzo Arboit, Giorgio Carlino, Jo\"el L. Lavanchy, Pietro Mascagni, Nicolas Padoy
Abstract: Acquiring surgical data for research and development is significantly hindered by high annotation costs and practical and ethical constraints. Utilizing synthetically generated images could offer a valuable alternative. In this work, we explore adapting text-to-image generative models for the surgical domain using the CholecT50 dataset, which provides surgical images annotated with action triplets (instrument, verb, target). We investigate several language models and find T5 to offer more distinct features for differentiating surgical actions on triplet-based textual inputs, and showcasing stronger alignment between long and triplet-based captions. To address challenges in training text-to-image models solely on triplet-based captions without additional inputs and supervisory signals, we discover that triplet text embeddings are instrument-centric in the latent space. Leveraging this insight, we design an instrument-based class balancing technique to counteract data imbalance and skewness, improving training convergence. Extending Imagen, a diffusion-based generative model, we develop Surgical Imagen to generate photorealistic and activity-aligned surgical images from triplet-based textual prompts. We assess the model on quality, alignment, reasoning, and knowledge, achieving FID and CLIP scores of 3.7 and 26.8% respectively. Human expert survey shows that participants were highly challenged by the realistic characteristics of the generated samples, demonstrating Surgical Imagen's effectiveness as a practical alternative to real data collection.
Authors: Yukai Shi, Zhipeng Weng, Yupei Lin, Cidan Shi, Xiaojun Yang, Liang Lin
Abstract: In recent years, deep neural networks tasks have increasingly relied on high-quality image inputs. With the development of high-resolution representation learning, the task of image dehazing has received significant attention. Previously, many methods collect diverse image data for large-scale training to boost the performance on a target scene. Ignoring the domain gap between different data, former de-hazing methods simply adopt multiple datasets for explicit large-scale training, which often makes the methods themselves be violated. To address this problem, we propose a novel method of cross-data vision alignment for richer representation learning to improve the existing dehazing methodology. Specifically, we call for the internal- and external knowledge should be further adapted with a self-supervised manner to fill up the domain gap. By using cross-data external alignment, the datasets inherit samples from different domains that are firmly aligned, making the model learn more robust and generalizable features. By using the internal augmentation method, the model can fully exploit local information within the images, and then obtaining more image details. To demonstrate the effectiveness of our proposed method, we conduct training on the Natural Image Dataset (NID). Experimental results show that our method clearly resolves the domain gap in different dehazing datasets and presents a new pipeline for large-scale training in the dehazing task. Our approach significantly outperforms other advanced methods in dehazing and produces dehazed images that are closest to real haze-free images.
Authors: Chun Tong Lei, Hon Ming Yam, Zhongliang Guo, Yifei Qian, Chun Pong Lau
Abstract: Neural networks have revolutionized numerous fields with their exceptional performance, yet they remain susceptible to adversarial attacks through subtle perturbations. While diffusion-based purification methods like DiffPure offer promising defense mechanisms, their computational overhead presents a significant practical limitation. In this paper, we introduce One Step Control Purification (OSCP), a novel defense framework that achieves robust adversarial purification in a single Neural Function Evaluation (NFE) within diffusion models. We propose Gaussian Adversarial Noise Distillation (GAND) as the distillation objective and Controlled Adversarial Purification (CAP) as the inference pipeline, which makes OSCP demonstrate remarkable efficiency while maintaining defense efficacy. Our proposed GAND addresses a fundamental tension between consistency distillation and adversarial perturbation, bridging the gap between natural and adversarial manifolds in the latent space, while remaining computationally efficient through Parameter-Efficient Fine-Tuning (PEFT) methods such as LoRA, eliminating the high computational budget request from full parameter fine-tuning. The CAP guides the purification process through the unlearnable edge detection operator calculated by the input image as an extra prompt, effectively preventing the purified images from deviating from their original appearance when large purification steps are used. Our experimental results on ImageNet showcase OSCP's superior performance, achieving a 74.19% defense success rate with merely 0.1s per purification -- a 100-fold speedup compared to conventional approaches.
Authors: Mengchen Zhang, Tong Wu, Tai Wang, Tengfei Wang, Ziwei Liu, Dahua Lin
Abstract: 6D object pose estimation aims at determining an object's translation, rotation, and scale, typically from a single RGBD image. Recent advancements have expanded this estimation from instance-level to category-level, allowing models to generalize across unseen instances within the same category. However, this generalization is limited by the narrow range of categories covered by existing datasets, such as NOCS, which also tend to overlook common real-world challenges like occlusion. To tackle these challenges, we introduce Omni6D, a comprehensive RGBD dataset featuring a wide range of categories and varied backgrounds, elevating the task to a more realistic context. 1) The dataset comprises an extensive spectrum of 166 categories, 4688 instances adjusted to the canonical pose, and over 0.8 million captures, significantly broadening the scope for evaluation. 2) We introduce a symmetry-aware metric and conduct systematic benchmarks of existing algorithms on Omni6D, offering a thorough exploration of new challenges and insights. 3) Additionally, we propose an effective fine-tuning approach that adapts models from previous datasets to our extensive vocabulary setting. We believe this initiative will pave the way for new insights and substantial progress in both the industrial and academic fields, pushing forward the boundaries of general 6D pose estimation.
Authors: Bohan Zhan, Wang Zhao, Yi Fang, Bo Du, Francisco Vasconcelos, Danail Stoyanov, Daniel S. Elson, Baoru Huang
Abstract: Accurate tracking of tissues and instruments in videos is crucial for Robotic-Assisted Minimally Invasive Surgery (RAMIS), as it enables the robot to comprehend the surgical scene with precise locations and interactions of tissues and tools. Traditional keypoint-based sparse tracking is limited by featured points, while flow-based dense two-view matching suffers from long-term drifts. Recently, the Tracking Any Point (TAP) algorithm was proposed to overcome these limitations and achieve dense accurate long-term tracking. However, its efficacy in surgical scenarios remains untested, largely due to the lack of a comprehensive surgical tracking dataset for evaluation. To address this gap, we introduce a new annotated surgical tracking dataset for benchmarking tracking methods for surgical scenarios, comprising real-world surgical videos with complex tissue and instrument motions. We extensively evaluate state-of-the-art (SOTA) TAP-based algorithms on this dataset and reveal their limitations in challenging surgical scenarios, including fast instrument motion, severe occlusions, and motion blur, etc. Furthermore, we propose a new tracking method, namely SurgMotion, to solve the challenges and further improve the tracking performance. Our proposed method outperforms most TAP-based algorithms in surgical instruments tracking, and especially demonstrates significant improvements over baselines in challenging medical videos. Our code and dataset are available at https://github.com/zhanbh1019/SurgicalMotion.
Authors: Jian Yang, Xukun Wang, Wentao Wang, Guoming Li, Qihang Fang, Ruihong Yuan, Tianyang Wang, Xiaomei Zhang, Yeying Jin, Zhaoxin Fan
Abstract: Audio-driven talking head generation is a pivotal area within film-making and Virtual Reality. Although existing methods have made significant strides following the end-to-end paradigm, they still encounter challenges in producing videos with high-frequency details due to their limited expressivity in this domain. This limitation has prompted us to explore an effective post-processing approach to synthesize photo-realistic talking head videos. Specifically, we employ a pretrained Wav2Lip model as our foundation model, leveraging its robust audio-lip alignment capabilities. Drawing on the theory of Lipschitz Continuity, we have theoretically established the noise robustness of Vector Quantised Auto Encoders (VQAEs). Our experiments further demonstrate that the high-frequency texture deficiency of the foundation model can be temporally consistently recovered by the Space-Optimised Vector Quantised Auto Encoder (SOVQAE) we introduced, thereby facilitating the creation of realistic talking head videos. We conduct experiments on both the conventional dataset and the High-Frequency TalKing head (HFTK) dataset that we curated. The results indicate that our method, LaDTalk, achieves new state-of-the-art video quality and out-of-domain lip synchronization performance.
Authors: Ahmed H. Salamah, Kaixiang Zheng, Yiwen Liu, En-Hui Yang
Abstract: Although it is traditionally believed that lossy image compression, such as JPEG compression, has a negative impact on the performance of deep neural networks (DNNs), it is shown by recent works that well-crafted JPEG compression can actually improve the performance of deep learning (DL). Inspired by this, we propose JPEG-DL, a novel DL framework that prepends any underlying DNN architecture with a trainable JPEG compression layer. To make the quantization operation in JPEG compression trainable, a new differentiable soft quantizer is employed at the JPEG layer, and then the quantization operation and underlying DNN are jointly trained. Extensive experiments show that in comparison with the standard DL, JPEG-DL delivers significant accuracy improvements across various datasets and model architectures while enhancing robustness against adversarial attacks. Particularly, on some fine-grained image classification datasets, JPEG-DL can increase prediction accuracy by as much as 20.9%. Our code is available on https://github.com/AhmedHussKhalifa/JPEG-Inspired-DL.git.
URLs: https://github.com/AhmedHussKhalifa/JPEG-Inspired-DL.git.
Authors: Saumya Gupta, Dimitris Samaras, Chao Chen
Abstract: Diffusion models excel at creating visually impressive images but often struggle to generate images with a specified topology. The Betti number, which represents the number of structures in an image, is a fundamental measure in topology. Yet, diffusion models fail to satisfy even this basic constraint. This limitation restricts their utility in applications requiring exact control, like robotics and environmental modeling. To address this, we propose TopoDiffusionNet (TDN), a novel approach that enforces diffusion models to maintain the desired topology. We leverage tools from topological data analysis, particularly persistent homology, to extract the topological structures within an image. We then design a topology-based objective function to guide the denoising process, preserving intended structures while suppressing noisy ones. Our experiments across four datasets demonstrate significant improvements in topological accuracy. TDN is the first to integrate topology with diffusion models, opening new avenues of research in this area. Code available at https://github.com/Saumya-Gupta-26/TopoDiffusionNet
Authors: Junjie Li, Jianghong Ma, Xiaofeng Zhang, Yuhang Li, Jianyang Shi
Abstract: Multimodal Large Language Models have advanced AI in applications like text-to-video generation and visual question answering. These models rely on visual encoders to convert non-text data into vectors, but current encoders either lack semantic alignment or overlook non-salient objects. We propose the Guiding Visual Encoder to Perceive Overlooked Information (GiVE) approach. GiVE enhances visual representation with an Attention-Guided Adapter (AG-Adapter) module and an Object-focused Visual Semantic Learning module. These incorporate three novel loss terms: Object-focused Image-Text Contrast (OITC) loss, Object-focused Image-Image Contrast (OIIC) loss, and Object-focused Image Discrimination (OID) loss, improving object consideration, retrieval accuracy, and comprehensiveness. Our contributions include dynamic visual focus adjustment, novel loss functions to enhance object retrieval, and the Multi-Object Instruction (MOInst) dataset. Experiments show our approach achieves state-of-the-art performance.
Authors: Yuxuan Lin, Yang Chang, Xuan Tong, Jiawen Yu, Antonio Liotta, Guofan Huang, Wei Song, Deyu Zeng, Zongze Wu, Yan Wang, Wenqiang Zhang
Abstract: In the advancement of industrial informatization, unsupervised anomaly detection technology effectively overcomes the scarcity of abnormal samples and significantly enhances the automation and reliability of smart manufacturing. As an important branch, industrial image anomaly detection focuses on automatically identifying visual anomalies in industrial scenarios (such as product surface defects, assembly errors, and equipment appearance anomalies) through computer vision techniques. With the rapid development of Unsupervised industrial Image Anomaly Detection (UIAD), excellent detection performance has been achieved not only in RGB setting but also in 3D and multimodal (RGB and 3D) settings. However, existing surveys primarily focus on UIAD tasks in RGB setting, with little discussion in 3D and multimodal settings. To address this gap, this artical provides a comprehensive review of UIAD tasks in the three modal settings. Specifically, we first introduce the task concept and process of UIAD. We then overview the research on UIAD in three modal settings (RGB, 3D, and multimodal), including datasets and methods, and review multimodal feature fusion strategies in multimodal setting. Finally, we summarize the main challenges faced by UIAD tasks in the three modal settings, and offer insights into future development directions, aiming to provide researchers with a comprehensive reference and offer new perspectives for the advancement of industrial informatization. Corresponding resources are available at https://github.com/Sunny5250/Awesome-Multi-Setting-UIAD.
URLs: https://github.com/Sunny5250/Awesome-Multi-Setting-UIAD.
Authors: Arash Marioriyad, Parham Rezaei, Mahdieh Soleymani Baghshah, Mohammad Hossein Rohban
Abstract: Text-to-image (T2I) generative models, such as Stable Diffusion and DALL-E, have shown remarkable proficiency in producing high-quality, realistic, and natural images from textual descriptions. However, these models sometimes fail to accurately capture all the details specified in the input prompts, particularly concerning entities, attributes, and spatial relationships. This issue becomes more pronounced when the prompt contains novel or complex compositions, leading to what are known as compositional generation failure modes. Recently, a new open-source diffusion-based T2I model, FLUX, has been introduced, demonstrating strong performance in high-quality image generation. Additionally, autoregressive T2I models like LlamaGen have claimed competitive visual quality performance compared to diffusion-based models. In this study, we evaluate the compositional generation capabilities of these newly introduced models against established models using the T2I-CompBench benchmark. Our findings reveal that LlamaGen, as a vanilla autoregressive model, is not yet on par with state-of-the-art diffusion models for compositional generation tasks under the same criteria, such as model size and inference time. On the other hand, the open-source diffusion-based model FLUX exhibits compositional generation capabilities comparable to the state-of-the-art closed-source model DALL-E3.
Authors: Shaohua Liu, Junzhe Lu, Zuoya Gu, Jiajun Li, Yue Deng
Abstract: Representing underwater 3D scenes is a valuable yet complex task, as attenuation and scattering effects during underwater imaging significantly couple the information of the objects and the water. This coupling presents a significant challenge for existing methods in effectively representing both the objects and the water medium simultaneously. To address this challenge, we propose Aquatic-GS, a hybrid 3D representation approach for underwater scenes that effectively represents both the objects and the water medium. Specifically, we construct a Neural Water Field (NWF) to implicitly model the water parameters, while extending the latest 3D Gaussian Splatting (3DGS) to model the objects explicitly. Both components are integrated through a physics-based underwater image formation model to represent complex underwater scenes. Moreover, to construct more precise scene geometry and details, we design a Depth-Guided Optimization (DGO) mechanism that uses a pseudo-depth map as auxiliary guidance. After optimization, Aquatic-GS enables the rendering of novel underwater viewpoints and supports restoring the true appearance of underwater scenes, as if the water medium were absent. Extensive experiments on both simulated and real-world datasets demonstrate that Aquatic-GS surpasses state-of-the-art underwater 3D representation methods, achieving better rendering quality and real-time rendering performance with a 410x increase in speed. Furthermore, regarding underwater image restoration, Aquatic-GS outperforms representative dewatering methods in color correction, detail recovery, and stability. Our models, code, and datasets can be accessed at https://aquaticgs.github.io.
Authors: Felix Tempel, Espen Alexander F. Ihlen, Lars Adde, Inga Str\"umke
Abstract: In Human Activity Recognition (HAR), understanding the intricacy of body movements within high-risk applications is essential. This study uses SHapley Additive exPlanations (SHAP) to explain the decision-making process of Graph Convolution Networks (GCNs) when classifying activities with skeleton data. We employ SHAP to explain two real-world datasets: one for cerebral palsy (CP) classification and the widely used NTU RGB+D 60 action recognition dataset. To test the explanation, we introduce a novel perturbation approach that modifies the model's edge importance matrix, allowing us to evaluate the impact of specific body key points on prediction outcomes. To assess the fidelity of our explanations, we employ informed perturbation, targeting body key points identified as important by SHAP and comparing them against random perturbation as a control condition. This perturbation enables a judgment on whether the body key points are truly influential or non-influential based on the SHAP values. Results on both datasets show that body key points identified as important through SHAP have the largest influence on the accuracy, specificity, and sensitivity metrics. Our findings highlight that SHAP can provide granular insights into the input feature contribution to the prediction outcome of GCNs in HAR tasks. This demonstrates the potential for more interpretable and trustworthy models in high-stakes applications like healthcare or rehabilitation.
Authors: Constantin Ulrich, Tassilo Wald, Emily Tempus, Maximilian Rokuss, Paul F. Jaeger, Klaus Maier-Hein
Abstract: Effortless and precise segmentation with minimal clinician effort could greatly streamline clinical workflows. Recent interactive segmentation models, inspired by METAs Segment Anything, have made significant progress but face critical limitations in 3D radiology. These include impractical human interaction requirements such as slice-by-slice operations for 2D models on 3D data and a lack of iterative refinement. Prior studies have been hindered by inadequate evaluation protocols, resulting in unreliable performance assessments and inconsistent findings across studies. The RadioActive benchmark addresses these challenges by providing a rigorous and reproducible evaluation framework for interactive segmentation methods in clinically relevant scenarios. It features diverse datasets, a wide range of target structures, and the most impactful 2D and 3D interactive segmentation methods, all within a flexible and extensible codebase. We also introduce advanced prompting techniques that reduce interaction steps, enabling fair comparisons between 2D and 3D models. Surprisingly, SAM2 outperforms all specialized medical 2D and 3D models in a setting requiring only a few interactions to generate prompts for a 3D volume. This challenges prevailing assumptions and demonstrates that general-purpose models surpass specialized medical approaches. By open-sourcing RadioActive, we invite researchers to integrate their models and prompting techniques, ensuring continuous and transparent evaluation of 3D medical interactive models.
Authors: Yongliang Wu, Xinting Hu, Yuyang Sun, Yizhou Zhou, Wenbo Zhu, Fengyun Rao, Bernt Schiele, Xu Yang
Abstract: Video Large Language Models (Vid-LLMs) have made remarkable advancements in comprehending video content for QA dialogue. However, they struggle to extend this visual understanding to tasks requiring precise temporal localization, known as Video Temporal Grounding (VTG). To address this gap, we introduce Number-Prompt (NumPro), a novel method that empowers Vid-LLMs to bridge visual comprehension with temporal grounding by adding unique numerical identifiers to each video frame. Treating a video as a sequence of numbered frame images, NumPro transforms VTG into an intuitive process: flipping through manga panels in sequence. This allows Vid-LLMs to "read" event timelines, accurately linking visual content with corresponding temporal information. Our experiments demonstrate that NumPro significantly boosts VTG performance of top-tier Vid-LLMs without additional computational cost. Furthermore, fine-tuning on a NumPro-enhanced dataset defines a new state-of-the-art for VTG, surpassing previous top-performing methods by up to 6.9\% in mIoU for moment retrieval and 8.5\% in mAP for highlight detection. The code will be available at https://github.com/yongliang-wu/NumPro.
Authors: Jiawei Zhang, Zijian Wu, Zhiyang Liang, Yicheng Gong, Dongfang Hu, Yao Yao, Xun Cao, Hao Zhu
Abstract: Reconstructing high-fidelity, animatable 3D head avatars from effortlessly captured monocular videos is a pivotal yet formidable challenge. Although significant progress has been made in rendering performance and manipulation capabilities, notable challenges remain, including incomplete reconstruction and inefficient Gaussian representation. To address these challenges, we introduce FATE, a novel method for reconstructing an editable full-head avatar from a single monocular video. FATE integrates a sampling-based densification strategy to ensure optimal positional distribution of points, improving rendering efficiency. A neural baking technique is introduced to convert discrete Gaussian representations into continuous attribute maps, facilitating intuitive appearance editing. Furthermore, we propose a universal completion framework to recover non-frontal appearance, culminating in a 360$^\circ$-renderable 3D head avatar. FATE outperforms previous approaches in both qualitative and quantitative evaluations, achieving state-of-the-art performance. To the best of our knowledge, FATE is the first animatable and 360$^\circ$ full-head monocular reconstruction method for a 3D head avatar.
Authors: Junho Kim, Hyunjun Kim, Hosu Lee, Yong Man Ro
Abstract: Despite advances in Large Multi-modal Models, applying them to long and untrimmed video content remains challenging due to limitations in context length and substantial memory overhead. These constraints often lead to significant information loss and reduced relevance in the model responses. With the exponential growth of video data across web platforms, understanding long-form video is crucial for advancing generalized intelligence. In this paper, we introduce SALOVA: Segment-Augmented LOng Video Assistant, a novel video-LLM framework designed to enhance the comprehension of lengthy video content through targeted retrieval process. We address two main challenges to achieve it: (i) We present the SceneWalk dataset, a high-quality collection of 87.8K long videos, each densely captioned at the segment level to enable models to capture scene continuity and maintain rich descriptive context. (ii) We develop robust architectural designs integrating dynamic routing mechanism and spatio-temporal projector to efficiently retrieve and process relevant video segments based on user queries. Our framework mitigates the limitations of current video-LMMs by allowing for precise identification and retrieval of relevant video segments in response to queries, thereby improving the contextual relevance of the generated responses. Through extensive experiments, SALOVA demonstrates enhanced capability in processing complex long-form videos, showing significant capability to maintain contextual integrity across extended sequences.
Authors: Zizhao Li, Zhengkang Xiang, Joseph West, Kourosh Khoshelham
Abstract: Traditional object detection methods operate under the closed-set assumption, where models can only detect a fixed number of objects predefined in the training set. Recent works on open vocabulary object detection (OVD) enable the detection of objects defined by an in-principle unbounded vocabulary, which reduces the cost of training models for specific tasks. However, OVD heavily relies on accurate prompts provided by an ``oracle'', which limits their use in critical applications such as driving scene perception. OVD models tend to misclassify near-out-of-distribution (NOOD) objects that have similar features to known classes, and ignore far-out-of-distribution (FOOD) objects. To address these limitations, we propose a framework that enables OVD models to operate in open world settings, by identifying and incrementally learning previously unseen objects. To detect FOOD objects, we propose Open World Embedding Learning (OWEL) and introduce the concept of Pseudo Unknown Embedding which infers the location of unknown classes in a continuous semantic space based on the information of known classes. We also propose Multi-Scale Contrastive Anchor Learning (MSCAL), which enables the identification of misclassified unknown objects by promoting the intra-class consistency of object embeddings at different scales. The proposed method achieves state-of-the-art performance on standard open world object detection and autonomous driving benchmarks while maintaining its open vocabulary object detection capability.
Authors: Wenxuan Huang, Zijie Zhai, Yunhang Shen, Shaosheng Cao, Fei Zhao, Xiangfeng Xu, Zheyu Ye, Yao Hu, Shaohui Lin
Abstract: Multimodal Large Language Models (MLLMs) have achieved remarkable success in vision understanding, reasoning, and interaction. However, the inference computation and memory increase progressively with the generation of output tokens during decoding, directly affecting the efficacy of MLLMs. Existing methods attempt to reduce the vision context redundancy to achieve efficient MLLMs. Unfortunately, the efficiency benefits of the vision context reduction in the prefill stage gradually diminish during the decoding stage. To address this problem, we proposed a dynamic vision-language context sparsification framework Dynamic-LLaVA, which dynamically reduces the redundancy of vision context in the prefill stage and decreases the memory and computation overhead of the generated language context during decoding. Dynamic-LLaVA designs a tailored sparsification inference scheme for different inference modes, i.e., prefill, decoding with and without KV cache, to achieve efficient inference of MLLMs. In practice, Dynamic-LLaVA can reduce computation consumption by $\sim$75\% in the prefill stage. Meanwhile, throughout the entire generation process of MLLMs, Dynamic-LLaVA reduces the $\sim$50\% computation consumption under decoding without KV cache, while saving $\sim$50\% GPU memory overhead when decoding with KV cache, due to the vision-language context sparsification. Extensive experiments also demonstrate that Dynamic-LLaVA achieves efficient inference for MLLMs with negligible understanding and generation ability degradation or even performance gains compared to the full-context inference baselines. Code is available at https://github.com/Osilly/dynamic_llava .
Authors: Chunlin Yu, Hanqing Wang, Ye Shi, Haoyang Luo, Sibei Yang, Jingyi Yu, Jingya Wang
Abstract: 3D affordance segmentation aims to link human instructions to touchable regions of 3D objects for embodied manipulations. Existing efforts typically adhere to single-object, single-affordance paradigms, where each affordance type or explicit instruction strictly corresponds to a specific affordance region and are unable to handle long-horizon tasks. Such a paradigm cannot actively reason about complex user intentions that often imply sequential affordances. In this paper, we introduce the Sequential 3D Affordance Reasoning task, which extends the traditional paradigm by reasoning from cumbersome user intentions and then decomposing them into a series of segmentation maps. Toward this, we construct the first instruction-based affordance segmentation benchmark that includes reasoning over both single and sequential affordances, comprising 180K instruction-point cloud pairs. Based on the benchmark, we propose our model, SeqAfford, to unlock the 3D multi-modal large language model with additional affordance segmentation abilities, which ensures reasoning with world knowledge and fine-grained affordance grounding in a cohesive framework. We further introduce a multi-granular language-point integration module to endow 3D dense prediction. Extensive experimental evaluations show that our model excels over well-established methods and exhibits open-world generalization with sequential reasoning abilities.
Authors: Ziwen Li, Jiaxin Huang, Runnan Chen, Yunlong Che, Yandong Guo, Tongliang Liu, Fakhri Karray, Mingming Gong
Abstract: Reconstructing urban scenes is challenging due to their complex geometries and the presence of potentially dynamic objects. 3D Gaussian Splatting (3DGS)-based methods have shown strong performance, but existing approaches often incorporate manual 3D annotations to improve dynamic object modeling, which is impractical due to high labeling costs. Some methods leverage 4D Gaussian Splatting (4DGS) to represent the entire scene, but they treat static and dynamic objects uniformly, leading to unnecessary updates for static elements and ultimately degrading reconstruction quality. To address these issues, we propose UrbanGS, which leverages 2D semantic maps and an existing dynamic Gaussian approach to distinguish static objects from the scene, enabling separate processing of definite static and potentially dynamic elements. Specifically, for definite static regions, we enforce global consistency to prevent unintended changes in dynamic Gaussian and introduce a K-nearest neighbor (KNN)-based regularization to improve local coherence on low-textured ground surfaces. Notably, for potentially dynamic objects, we aggregate temporal information using learnable time embeddings, allowing each Gaussian to model deformations over time. Extensive experiments on real-world datasets demonstrate that our approach outperforms state-of-the-art methods in reconstruction quality and efficiency, accurately preserving static content while capturing dynamic elements.
Authors: Seokju Yun, Seunghye Chae, Dongheon Lee, Youngmin Ro
Abstract: Domain generalization (DG) aims to adapt a model using one or multiple source domains to ensure robust performance in unseen target domains. Recently, Parameter-Efficient Fine-Tuning (PEFT) of foundation models has shown promising results in the context of DG problem. Nevertheless, existing PEFT methods still struggle to strike a balance between preserving generalizable components of the pre-trained model and learning task-specific features. To gain insights into the distribution of generalizable components, we begin by analyzing the pre-trained weights through the lens of singular value decomposition. Building on these insights, we introduce Singular Value Decomposed Minor Components Adaptation (SoMA), an approach that selectively tunes minor singular components while keeping the residual parts frozen. SoMA effectively retains the generalization ability of the pre-trained model while efficiently acquiring task-specific skills. Moreover, we freeze domain-generalizable blocks and employ an annealing weight decay strategy, thereby achieving an optimal balance in the delicate trade-off between generalizability and discriminability. SoMA attains state-of-the-art results on multiple benchmarks that span both domain generalized semantic segmentation to domain generalized object detection. In addition, our methods introduce no additional inference overhead or regularization loss, maintain compatibility with any backbone or head, and are designed to be versatile, allowing easy integration into a wide range of tasks.
Authors: Hyesu Lim, Jinho Choi, Jaegul Choo, Steffen Schneider
Abstract: Adapting foundation models for specific purposes has become a standard approach to build machine learning systems for downstream applications. Yet, it is an open question which mechanisms take place during adaptation. Here we develop a new Sparse Autoencoder (SAE) for the CLIP vision transformer, named PatchSAE, to extract interpretable concepts at granular levels (e.g., shape, color, or semantics of an object) and their patch-wise spatial attributions. We explore how these concepts influence the model output in downstream image classification tasks and investigate how recent state-of-the-art prompt-based adaptation techniques change the association of model inputs to these concepts. While activations of concepts slightly change between adapted and non-adapted models, we find that the majority of gains on common adaptation tasks can be explained with the existing concepts already present in the non-adapted foundation model. This work provides a concrete framework to train and use SAEs for Vision Transformers and provides insights into explaining adaptation mechanisms.
Authors: Jan Rathjens, Shirin Reyhanian, David Kappel, Laurenz Wiskott
Abstract: Understanding the mechanisms underlying deep neural networks in computer vision remains a fundamental challenge. While many previous approaches have focused on visualizing intermediate representations within deep neural networks, particularly convolutional neural networks, these techniques have yet to be thoroughly explored in transformer-based vision models. In this study, we apply a modular approach of training inverse models to reconstruct input images from intermediate layers within a Detection Transformer and a Vision Transformer, showing that this approach is efficient and feasible. Through qualitative and quantitative evaluations of reconstructed images, we generate insights into the underlying mechanisms of these architectures, highlighting their similarities and differences in terms of contextual shape and preservation of image details, inter-layer correlation, and robustness to color perturbations. Our analysis illustrates how these properties emerge within the models, contributing to a deeper understanding of transformer-based vision models. The code for reproducing our experiments is available at github.com/wiskott-lab/inverse-detection-transformer.
Authors: Baorui Ma, Huachen Gao, Haoge Deng, Zhengxiong Luo, Tiejun Huang, Lulu Tang, Xinlong Wang
Abstract: Recent 3D generation models typically rely on limited-scale 3D `gold-labels' or 2D diffusion priors for 3D content creation. However, their performance is upper-bounded by constrained 3D priors due to the lack of scalable learning paradigms. In this work, we present See3D, a visual-conditional multi-view diffusion model trained on large-scale Internet videos for open-world 3D creation. The model aims to Get 3D knowledge by solely Seeing the visual contents from the vast and rapidly growing video data -- You See it, You Got it. To achieve this, we first scale up the training data using a proposed data curation pipeline that automatically filters out multi-view inconsistencies and insufficient observations from source videos. This results in a high-quality, richly diverse, large-scale dataset of multi-view images, termed WebVi3D, containing 320M frames from 16M video clips. Nevertheless, learning generic 3D priors from videos without explicit 3D geometry or camera pose annotations is nontrivial, and annotating poses for web-scale videos is prohibitively expensive. To eliminate the need for pose conditions, we introduce an innovative visual-condition - a purely 2D-inductive visual signal generated by adding time-dependent noise to the masked video data. Finally, we introduce a novel visual-conditional 3D generation framework by integrating See3D into a warping-based pipeline for high-fidelity 3D generation. Our numerical and visual comparisons on single and sparse reconstruction benchmarks show that See3D, trained on cost-effective and scalable video data, achieves notable zero-shot and open-world generation capabilities, markedly outperforming models trained on costly and constrained 3D datasets. Please refer to our project page at: https://vision.baai.ac.cn/see3d
Authors: Siyan Dong, Shuzhe Wang, Shaohui Liu, Lulu Cai, Qingnan Fan, Juho Kannala, Yanchao Yang
Abstract: Visual localization aims to determine the camera pose of a query image relative to a database of posed images. In recent years, deep neural networks that directly regress camera poses have gained popularity due to their fast inference capabilities. However, existing methods struggle to either generalize well to new scenes or provide accurate camera pose estimates. To address these issues, we present Reloc3r, a simple yet effective visual localization framework. It consists of an elegantly designed relative pose regression network, and a minimalist motion averaging module for absolute pose estimation. Trained on approximately eight million posed image pairs, Reloc3r achieves surprisingly good performance and generalization ability. We conduct extensive experiments on six public datasets, consistently demonstrating the effectiveness and efficiency of the proposed method. It provides high-quality camera pose estimates in real time and generalizes to novel scenes. Code: https://github.com/ffrivera0/reloc3r.
Authors: Siyuan Wang, Dianyi Wang, Chengxing Zhou, Zejun Li, Zhihao Fan, Xuanjing Huang, Zhongyu Wei
Abstract: Large Vision-Language Models (LVLMs) typically learn visual capacity through visual instruction tuning, involving updates to both a projector and their LLM backbones. Inspired by the concept of a visual region in the human brain, we investigate the existence of an analogous \textit{visual region} within LLMs that functions as a cognitive core, and explore the potential of efficient training of LVLMs via selective layers tuning. Using Bunny-Llama-3-8B-V for detailed analysis and other three LVLMs for validation across diverse visual and textual tasks, we find that selectively updating 25\% of LLMs layers, when sparsely and uniformly distributed, can preserve nearly 99\% of visual performance and maintain or improve textual task results, while effectively reducing training time. Based on this targeted training approach, we further propose a novel visual region-based pruning paradigm, removing non-critical layers outside the visual region, which can achieve minimal performance loss. This study offers an effective and efficient strategy for LVLM training and inference by activating a layer-wise visual region within LLMs, which proves consistently effective across different models.
Authors: Hongjie Li, Hong-Xing Yu, Jiaman Li, Jiajun Wu
Abstract: Human-scene interaction (HSI) generation is crucial for applications in embodied AI, virtual reality, and robotics. Yet, existing methods cannot synthesize interactions in unseen environments such as in-the-wild scenes or reconstructed scenes, as they rely on paired 3D scenes and captured human motion data for training, which are unavailable for unseen environments. We present ZeroHSI, a novel approach that enables zero-shot 4D human-scene interaction synthesis, eliminating the need for training on any MoCap data. Our key insight is to distill human-scene interactions from state-of-the-art video generation models, which have been trained on vast amounts of natural human movements and interactions, and use differentiable rendering to reconstruct human-scene interactions. ZeroHSI can synthesize realistic human motions in both static scenes and environments with dynamic objects, without requiring any ground-truth motion data. We evaluate ZeroHSI on a curated dataset of different types of various indoor and outdoor scenes with different interaction prompts, demonstrating its ability to generate diverse and contextually appropriate human-scene interactions.
Authors: Shiyao Li, Yingchun Hu, Xuefei Ning, Xihui Liu, Ke Hong, Xiaotao Jia, Xiuhong Li, Yaqi Yan, Pei Ran, Guohao Dai, Shengen Yan, Huazhong Yang, Yu Wang
Abstract: Vision-Language Models (VLMs) have enabled a variety of real-world applications. The large parameter size of VLMs brings large memory and computation overhead which poses significant challenges for deployment. Post-Training Quantization (PTQ) is an effective technique to reduce the memory and computation overhead. Existing PTQ methods mainly focus on large language models (LLMs), without considering the differences across other modalities. In this paper, we discover that there is a significant difference in sensitivity between language and vision tokens in large VLMs. Therefore, treating tokens from different modalities equally, as in existing PTQ methods, may over-emphasize the insensitive modalities, leading to significant accuracy loss. To deal with the above issue, we propose a simple yet effective method, Modality-Balanced Quantization (MBQ), for large VLMs. Specifically, MBQ incorporates the different sensitivities across modalities during the calibration process to minimize the reconstruction loss for better quantization parameters. Extensive experiments show that MBQ can significantly improve task accuracy by up to 4.4% and 11.6% under W3 and W4A8 quantization for 7B to 70B VLMs, compared to SOTA baselines. Additionally, we implement a W3 GPU kernel that fuses the dequantization and GEMV operators, achieving a 1.4x speedup on LLaVA-onevision-7B on the RTX 4090. The code is available at https://github.com/thu-nics/MBQ.
Authors: Olivier Morelle (B-IT and Department of Computer Science, University of Bonn, Department of Ophthalmology, University Hospital Bonn), Justus Bisten (B-IT and Department of Computer Science, University of Bonn), Maximilian W. M. Wintergerst (Department of Ophthalmology, University Hospital Bonn, Augenzentrum Grischun, Chur, Switzerland), Robert P. Finger (Department of Ophthalmology, University Hospital Bonn, Department of Ophthalmology, University Medical Center Mannheim, Heidelberg University), Thomas Schultz (B-IT and Department of Computer Science, University of Bonn, Lamarr Institute for Machine Learning and Artificial Intelligence)
Abstract: Weakly supervised segmentation has the potential to greatly reduce the annotation effort for training segmentation models for small structures such as hyper-reflective foci (HRF) in optical coherence tomography (OCT). However, most weakly supervised methods either involve a strong downsampling of input images, or only achieve localization at a coarse resolution, both of which are unsatisfactory for small structures. We propose a novel framework that increases the spatial resolution of a traditional attention-based Multiple Instance Learning (MIL) approach by using Layer-wise Relevance Propagation (LRP) to prompt the Segment Anything Model (SAM~2), and increases recall with iterative inference. Moreover, we demonstrate that replacing MIL with a Compact Convolutional Transformer (CCT), which adds a positional encoding, and permits an exchange of information between different regions of the OCT image, leads to a further and substantial increase in segmentation accuracy.
Authors: Xianwei Zhuang, Zhihong Zhu, Yuxin Xie, Liming Liang, Yuexian Zou
Abstract: Large Vision-Language Models (LVLMs) may produce outputs that are unfaithful to reality, also known as visual hallucinations (VH), which significantly impedes their real-world usage. To alleviate VH, various decoding strategies have been proposed to enhance visual information. However, many of these methods may require secondary decoding and rollback, which significantly reduces inference speed. In this work, we propose an efficient plug-and-play decoding algorithm via Visual-Aware Sparsification (VASparse) from the perspective of token sparsity for mitigating VH. VASparse is inspired by empirical observations: (1) the sparse activation of attention in LVLMs, and (2) visual-agnostic tokens sparsification exacerbates VH. Based on these insights, we propose a novel token sparsification strategy that balances efficiency and trustworthiness. Specifically, VASparse implements a visual-aware token selection strategy during decoding to reduce redundant tokens while preserving visual context effectively. Additionally, we innovatively introduce a sparse-based visual contrastive decoding method to recalibrate the distribution of hallucinated outputs without the time overhead associated with secondary decoding. Subsequently, VASparse recalibrates attention scores to penalize attention sinking of LVLMs towards text tokens. Extensive experiments across four popular benchmarks confirm the effectiveness of VASparse in mitigating VH across different LVLM families without requiring additional training or post-processing. Impressively, VASparse achieves state-of-the-art performance for mitigating VH while maintaining competitive decoding speed. Code is available at https://github.com/mengchuang123/VASparse-github.
Authors: Olga Zatsarynna, Emad Bahrami, Yazan Abu Farha, Gianpiero Francesca, Juergen Gall
Abstract: Long-term dense action anticipation is very challenging since it requires predicting actions and their durations several minutes into the future based on provided video observations. To model the uncertainty of future outcomes, stochastic models predict several potential future action sequences for the same observation. Recent work has further proposed to incorporate uncertainty modelling for observed frames by simultaneously predicting per-frame past and future actions in a unified manner. While such joint modelling of actions is beneficial, it requires long-range temporal capabilities to connect events across distant past and future time points. However, the previous work struggles to achieve such a long-range understanding due to its limited and/or sparse receptive field. To alleviate this issue, we propose a novel MANTA (MAmba for ANTicipation) network. Our model enables effective long-term temporal modelling even for very long sequences while maintaining linear complexity in sequence length. We demonstrate that our approach achieves state-of-the-art results on three datasets - Breakfast, 50Salads, and Assembly101 - while also significantly improving computational and memory efficiency. Our code is available at https://github.com/olga-zats/DIFF_MANTA .
Authors: Junyu Xia, Jiesong Bai, Yihang Dong
Abstract: Low-light image enhancement (LLE) aims to improve the visual quality of images captured in poorly lit conditions, which often suffer from low brightness, low contrast, noise, and color distortions. These issues hinder the performance of computer vision tasks such as object detection, facial recognition, and autonomous driving.Traditional enhancement techniques, such as multi-scale fusion and histogram equalization, fail to preserve fine details and often struggle with maintaining the natural appearance of enhanced images under complex lighting conditions. Although the Retinex theory provides a foundation for image decomposition, it often amplifies noise, leading to suboptimal image quality. In this paper, we propose the Dual Light Enhance Network (DLEN), a novel architecture that incorporates two distinct attention mechanisms, considering both spatial and frequency domains. Our model introduces a learnable wavelet transform module in the illumination estimation phase, preserving high- and low-frequency components to enhance edge and texture details. Additionally, we design a dual-branch structure that leverages the power of the Transformer architecture to enhance both the illumination and structural components of the image.Through extensive experiments, our model outperforms state-of-the-art methods on standard benchmarks.Code is available here: https://github.com/LaLaLoXX/DLEN
Authors: Aashish Rai, Dilin Wang, Mihir Jain, Nikolaos Sarafianos, Kefan Chen, Srinath Sridhar, Aayush Prakash
Abstract: 3D Gaussian Splatting (3DGS) has demonstrated superior quality in modeling 3D objects and scenes. However, generating 3DGS remains challenging due to their discrete, unstructured, and permutation-invariant nature. In this work, we present a simple yet effective method to overcome these challenges. We utilize spherical mapping to transform 3DGS into a structured 2D representation, termed UVGS. UVGS can be viewed as multi-channel images, with feature dimensions as a concatenation of Gaussian attributes such as position, scale, color, opacity, and rotation. We further find that these heterogeneous features can be compressed into a lower-dimensional (e.g., 3-channel) shared feature space using a carefully designed multi-branch network. The compressed UVGS can be treated as typical RGB images. Remarkably, we discover that typical VAEs trained with latent diffusion models can directly generalize to this new representation without additional training. Our novel representation makes it effortless to leverage foundational 2D models, such as diffusion models, to directly model 3DGS. Additionally, one can simply increase the 2D UV resolution to accommodate more Gaussians, making UVGS a scalable solution compared to typical 3D backbones. This approach immediately unlocks various novel generation applications of 3DGS by inherently utilizing the already developed superior 2D generation capabilities. In our experiments, we demonstrate various unconditional, conditional generation, and inpainting applications of 3DGS based on diffusion models, which were previously non-trivial.
Authors: Quan Zhang, Yuxin Qi, Xi Tang, Jinwei Fang, Xi Lin, Ke Zhang, Chun Yuan
Abstract: Using extensive training data from SA-1B, the Segment Anything Model (SAM) has demonstrated exceptional generalization and zero-shot capabilities, attracting widespread attention in areas such as medical image segmentation and remote sensing image segmentation. However, its performance in the field of image manipulation detection remains largely unexplored and unconfirmed. There are two main challenges in applying SAM to image manipulation detection: a) reliance on manual prompts, and b) the difficulty of single-view information in supporting cross-dataset generalization. To address these challenges, we develops a cross-view prompt learning paradigm called IMDPrompter based on SAM. Benefiting from the design of automated prompts, IMDPrompter no longer relies on manual guidance, enabling automated detection and localization. Additionally, we propose components such as Cross-view Feature Perception, Optimal Prompt Selection, and Cross-View Prompt Consistency, which facilitate cross-view perceptual learning and guide SAM to generate accurate masks. Extensive experimental results from five datasets (CASIA, Columbia, Coverage, IMD2020, and NIST16) validate the effectiveness of our proposed method.
Authors: Sheng Zhou, Junbin Xiao, Qingyun Li, Yicong Li, Xun Yang, Dan Guo, Meng Wang, Tat-Seng Chua, Angela Yao
Abstract: We introduce EgoTextVQA, a novel and rigorously constructed benchmark for egocentric QA assistance involving scene text. EgoTextVQA contains 1.5K ego-view videos and 7K scene-text aware questions that reflect real user needs in outdoor driving and indoor house-keeping activities. The questions are designed to elicit identification and reasoning on scene text in an egocentric and dynamic environment. With EgoTextVQA, we comprehensively evaluate 10 prominent multimodal large language models. Currently, all models struggle, and the best results (Gemini 1.5 Pro) are around 33\% accuracy, highlighting the severe deficiency of these techniques in egocentric QA assistance. Our further investigations suggest that precise temporal grounding and multi-frame reasoning, along with high resolution and auxiliary scene-text inputs, are key for better performance. With thorough analyses and heuristic suggestions, we hope EgoTextVQA can serve as a solid testbed for research in egocentric scene-text QA assistance. Our dataset is released at: https://github.com/zhousheng97/EgoTextVQA.
Authors: Yuanxun Lu, Jingyang Zhang, Tian Fang, Jean-Daniel Nahmias, Yanghai Tsin, Long Quan, Xun Cao, Yao Yao, Shiwei Li
Abstract: We present Matrix3D, a unified model that performs several photogrammetry subtasks, including pose estimation, depth prediction, and novel view synthesis using just the same model. Matrix3D utilizes a multi-modal diffusion transformer (DiT) to integrate transformations across several modalities, such as images, camera parameters, and depth maps. The key to Matrix3D's large-scale multi-modal training lies in the incorporation of a mask learning strategy. This enables full-modality model training even with partially complete data, such as bi-modality data of image-pose and image-depth pairs, thus significantly increases the pool of available training data. Matrix3D demonstrates state-of-the-art performance in pose estimation and novel view synthesis tasks. Additionally, it offers fine-grained control through multi-round interactions, making it an innovative tool for 3D content creation. Project page: https://nju-3dv.github.io/projects/matrix3d.
Authors: Gamal Elghazaly, Raphael Frank
Abstract: High-Definition (HD) maps play a crucial role in autonomous vehicle navigation, complementing onboard perception sensors for improved accuracy and safety. Traditional HD map generation relies on dedicated mapping vehicles, which are costly and fail to capture real-time infrastructure changes. This paper presents HDMapLaneNet, a novel framework leveraging V2X communication and Scene Graph Generation to collaboratively construct a localized geometric layer of HD maps. The approach extracts lane centerlines from front-facing camera images, represents them as graphs, and transmits the data for global aggregation to the cloud via V2X. Preliminary results on the nuScenes dataset demonstrate superior association prediction performance compared to a state-of-the-art method.
Authors: Yinghao Shuai, Ran Yu, Yuantao Chen, Zijian Jiang, Xiaowei Song, Nan Wang, Jv Zheng, Jianzhu Ma, Meng Yang, Zhicheng Wang, Wenbo Ding, Hao Zhao
Abstract: Current robotic systems can understand the categories and poses of objects well. But understanding physical properties like mass, friction, and hardness, in the wild, remains challenging. We propose a new method that reconstructs 3D objects using the Gaussian splatting representation and predicts various physical properties in a zero-shot manner. We propose two techniques during the reconstruction phase: a geometry-aware regularization loss function to improve the shape quality and a region-aware feature contrastive loss function to promote region affinity. Two other new techniques are designed during inference: a feature-based property propagation module and a volume integration module tailored for the Gaussian representation. Our framework is named as zero-shot physical understanding with Gaussian splatting, or PUGS. PUGS achieves new state-of-the-art results on the standard benchmark of ABO-500 mass prediction. We provide extensive quantitative ablations and qualitative visualization to demonstrate the mechanism of our designs. We show the proposed methodology can help address challenging real-world grasping tasks. Our codes, data, and models are available at https://github.com/EverNorif/PUGS
Authors: Farzad Beizaee, Gregory Lodygensky, Christian Desrosiers, Jose Dolz
Abstract: Unsupervised anomaly detection in brain images is crucial for identifying injuries and pathologies without access to labels. However, the accurate localization of anomalies in medical images remains challenging due to the inherent complexity and variability of brain structures and the scarcity of annotated abnormal data. To address this challenge, we propose a novel approach that incorporates masking within diffusion models, leveraging their generative capabilities to learn robust representations of normal brain anatomy. During training, our model processes only normal brain MRI scans and performs a forward diffusion process in the latent space that adds noise to the features of randomly-selected patches. Following a dual objective, the model learns to identify which patches are noisy and recover their original features. This strategy ensures that the model captures intricate patterns of normal brain structures while isolating potential anomalies as noise in the latent space. At inference, the model identifies noisy patches corresponding to anomalies and generates a normal counterpart for these patches by applying a reverse diffusion process. Our method surpasses existing unsupervised anomaly detection techniques, demonstrating superior performance in generating accurate normal counterparts and localizing anomalies. The code is available at hhttps://github.com/farzad-bz/MAD-AD.
Authors: Sucheng Ren, Qihang Yu, Ju He, Xiaohui Shen, Alan Yuille, Liang-Chieh Chen
Abstract: Autoregressive (AR) modeling, known for its next-token prediction paradigm, underpins state-of-the-art language and visual generative models. Traditionally, a ``token'' is treated as the smallest prediction unit, often a discrete symbol in language or a quantized patch in vision. However, the optimal token definition for 2D image structures remains an open question. Moreover, AR models suffer from exposure bias, where teacher forcing during training leads to error accumulation at inference. In this paper, we propose xAR, a generalized AR framework that extends the notion of a token to an entity X, which can represent an individual patch token, a cell (a $k\times k$ grouping of neighboring patches), a subsample (a non-local grouping of distant patches), a scale (coarse-to-fine resolution), or even a whole image. Additionally, we reformulate discrete token classification as continuous entity regression, leveraging flow-matching methods at each AR step. This approach conditions training on noisy entities instead of ground truth tokens, leading to Noisy Context Learning, which effectively alleviates exposure bias. As a result, xAR offers two key advantages: (1) it enables flexible prediction units that capture different contextual granularity and spatial structures, and (2) it mitigates exposure bias by avoiding reliance on teacher forcing. On ImageNet-256 generation benchmark, our base model, xAR-B (172M), outperforms DiT-XL/SiT-XL (675M) while achieving 20$\times$ faster inference. Meanwhile, xAR-H sets a new state-of-the-art with an FID of 1.24, running 2.2$\times$ faster than the previous best-performing model without relying on vision foundation modules (e.g., DINOv2) or advanced guidance interval sampling.
Authors: Yifei Qian, Zhongliang Guo, Bowen Deng, Chun Tong Lei, Shuai Zhao, Chun Pong Lau, Xiaopeng Hong, Michael P. Pound
Abstract: Zero-shot object counting aims to count instances of arbitrary object categories specified by text descriptions. Existing methods typically rely on vision-language models like CLIP, but often exhibit limited sensitivity to text prompts. We present T2ICount, a diffusion-based framework that leverages rich prior knowledge and fine-grained visual understanding from pretrained diffusion models. While one-step denoising ensures efficiency, it leads to weakened text sensitivity. To address this challenge, we propose a Hierarchical Semantic Correction Module that progressively refines text-image feature alignment, and a Representational Regional Coherence Loss that provides reliable supervision signals by leveraging the cross-attention maps extracted from the denosing U-Net. Furthermore, we observe that current benchmarks mainly focus on majority objects in images, potentially masking models' text sensitivity. To address this, we contribute a challenging re-annotated subset of FSC147 for better evaluation of text-guided counting ability. Extensive experiments demonstrate that our method achieves superior performance across different benchmarks. Code is available at https://github.com/cha15yq/T2ICount.
Authors: Ziyue Huang, Yongchao Feng, Shuai Yang, Ziqi Liu, Qingjie Liu, Yunhong Wang
Abstract: Remote sensing object detection has made significant progress, but most studies still focus on closed-set detection, limiting generalization across diverse datasets. Open-vocabulary object detection (OVD) provides a solution by leveraging multimodal associations between text prompts and visual features. However, existing OVD methods for remote sensing (RS) images are constrained by small-scale datasets and fail to address the unique challenges of remote sensing interpretation, include oriented object detection and the need for both high precision and real-time performance in diverse scenarios. To tackle these challenges, we propose OpenRSD, a universal open-prompt RS object detection framework. OpenRSD supports multimodal prompts and integrates multi-task detection heads to balance accuracy and real-time requirements. Additionally, we design a multi-stage training pipeline to enhance the generalization of model. Evaluated on seven public datasets, OpenRSD demonstrates superior performance in oriented and horizontal bounding box detection, with real-time inference capabilities suitable for large-scale RS image analysis. Compared to YOLO-World, OpenRSD exhibits an 8.7\% higher average precision and achieves an inference speed of 20.8 FPS. Codes and models will be released.
Authors: Sahar Dastani, Ali Bahri, Moslem Yazdanpanah, Mehrdad Noori, David Osowiechi, Gustavo Adolfo Vargas Hakim, Farzad Beizaee, Milad Cheraghalikhani, Arnab Kumar Mondal, Herve Lombaert, Christian Desrosiers
Abstract: State Space Models (SSMs) have recently emerged as an alternative to Vision Transformers (ViTs) due to their unique ability of modeling global relationships with linear complexity. SSMs are specifically designed to capture spatially proximate relationships of image patches. However, they fail to identify relationships between conceptually related yet not adjacent patches. This limitation arises from the non-causal nature of image data, which lacks inherent directional relationships. Additionally, current vision-based SSMs are highly sensitive to transformations such as rotation. Their predefined scanning directions depend on the original image orientation, which can cause the model to produce inconsistent patch-processing sequences after rotation. To address these limitations, we introduce Spectral VMamba, a novel approach that effectively captures the global structure within an image by leveraging spectral information derived from the graph Laplacian of image patches. Through spectral decomposition, our approach encodes patch relationships independently of image orientation, achieving rotation invariance with the aid of our Rotational Feature Normalizer (RFN) module. Our experiments on classification tasks show that Spectral VMamba outperforms the leading SSM models in vision, such as VMamba, while maintaining invariance to rotations and a providing a similar runtime efficiency.
Authors: Chenfei Liao, Xu Zheng, Yuanhuiyi Lyu, Haiwei Xue, Yihong Cao, Jiawen Wang, Kailun Yang, Xuming Hu
Abstract: Research has focused on Multi-Modal Semantic Segmentation (MMSS), where pixel-wise predictions are derived from multiple visual modalities captured by diverse sensors. Recently, the large vision model, Segment Anything Model 2 (SAM2), has shown strong zero-shot segmentation performance on both images and videos. When extending SAM2 to MMSS, two issues arise: 1. How can SAM2 be adapted to multi-modal data? 2. How can SAM2 better understand semantics? Inspired by cross-frame correlation in videos, we propose to treat multi-modal data as a sequence of frames representing the same scene. Our key idea is to ''memorize'' the modality-agnostic information and 'memorize' the semantics related to the targeted scene. To achieve this, we apply SAM2's memory mechanisms across multi-modal data to capture modality-agnostic features. Meanwhile, to memorize the semantic knowledge, we propose a training-only Semantic Prototype Memory Module (SPMM) to store category-level prototypes across training for facilitating SAM2's transition from instance to semantic segmentation. A prototypical adaptation loss is imposed between global and local prototypes iteratively to align and refine SAM2's semantic understanding. Extensive experimental results demonstrate that our proposed MemorySAM outperforms SoTA methods by large margins on both synthetic and real-world benchmarks (65.38% on DELIVER, 52.88% on MCubeS). Source code will be made publicly available.
Authors: Hanqing Liu, Shouwei Ruan, Yao Huang, Shiji Zhao, Xingxing Wei
Abstract: Vision-Language Models (VLMs) have achieved remarkable success in various tasks, yet their robustness to real-world illumination variations remains largely unexplored. To bridge this gap, we propose \textbf{I}llumination \textbf{T}ransformation \textbf{A}ttack (\textbf{ITA}), the first framework to systematically assess VLMs' robustness against illumination changes. However, there still exist two key challenges: (1) how to model global illumination with fine-grained control to achieve diverse lighting conditions and (2) how to ensure adversarial effectiveness while maintaining naturalness. To address the first challenge, we innovatively decompose global illumination into multiple parameterized point light sources based on the illumination rendering equation. This design enables us to model more diverse lighting variations that previous methods could not capture. Then, by integrating these parameterized lighting variations with physics-based lighting reconstruction techniques, we could precisely render such light interactions in the original scenes, finally meeting the goal of fine-grained lighting control. For the second challenge, by controlling illumination through the lighting reconstrution model's latent space rather than direct pixel manipulation, we inherently preserve physical lighting priors. Furthermore, to prevent potential reconstruction artifacts, we design additional perceptual constraints for maintaining visual consistency with original images and diversity constraints for avoiding light source convergence. Extensive experiments demonstrate that our ITA could significantly reduce the performance of advanced VLMs, e.g., LLaVA-1.6, while possessing competitive naturalness, exposing VLMS' critical illuminiation vulnerabilities.
Authors: Boeun Kim, Hea In Jeong, JungHoon Sung, Yihua Cheng, Jeongmin Lee, Ju Yong Chang, Sang-Il Choi, Younggeun Choi, Saim Shin, Jungho Kim, Hyung Jin Chang
Abstract: This paper introduces Motion Personalization, a new task that generates personalized motions aligned with text descriptions using several basic motions containing Persona. To support this novel task, we introduce a new large-scale motion dataset called PerMo (PersonaMotion), which captures the unique personas of multiple actors. We also propose a multi-modal finetuning method of a pretrained motion diffusion model called PersonaBooth. PersonaBooth addresses two main challenges: i) A significant distribution gap between the persona-focused PerMo dataset and the pretraining datasets, which lack persona-specific data, and ii) the difficulty of capturing a consistent persona from the motions vary in content (action type). To tackle the dataset distribution gap, we introduce a persona token to accept new persona features and perform multi-modal adaptation for both text and visuals during finetuning. To capture a consistent persona, we incorporate a contrastive learning technique to enhance intra-cohesion among samples with the same persona. Furthermore, we introduce a context-aware fusion mechanism to maximize the integration of persona cues from multiple input motions. PersonaBooth outperforms state-of-the-art motion style transfer methods, establishing a new benchmark for motion personalization.
Authors: Ethan Griffiths, Maryam Haghighat, Simon Denman, Clinton Fookes, Milad Ramezani
Abstract: We present HOTFormerLoc, a novel and versatile Hierarchical Octree-based TransFormer, for large-scale 3D place recognition in both ground-to-ground and ground-to-aerial scenarios across urban and forest environments. We propose an octree-based multi-scale attention mechanism that captures spatial and semantic features across granularities. To address the variable density of point distributions from spinning lidar, we present cylindrical octree attention windows to reflect the underlying distribution during attention. We introduce relay tokens to enable efficient global-local interactions and multi-scale representation learning at reduced computational cost. Our pyramid attentional pooling then synthesises a robust global descriptor for end-to-end place recognition in challenging environments. In addition, we introduce CS-Wild-Places, a novel 3D cross-source dataset featuring point cloud data from aerial and ground lidar scans captured in dense forests. Point clouds in CS-Wild-Places contain representational gaps and distinctive attributes such as varying point densities and noise patterns, making it a challenging benchmark for cross-view localisation in the wild. HOTFormerLoc achieves a top-1 average recall improvement of 5.5% - 11.5% on the CS-Wild-Places benchmark. Furthermore, it consistently outperforms SOTA 3D place recognition methods, with an average performance gain of 4.9% on well-established urban and forest datasets. The code and CS-Wild-Places benchmark is available at https://csiro-robotics.github.io/HOTFormerLoc.
Authors: Jaa-Yeon Lee, Byunghee Cha, Jeongsol Kim, Jong Chul Ye
Abstract: While recent advancements in generative modeling have significantly improved text-image alignment, some residual misalignment between text and image representations still remains. Although many approaches have attempted to address this issue by fine-tuning models using various reward models, etc., we revisit the challenge from the perspective of representation alignment-an approach that has gained popularity with the success of REPresentation Alignment (REPA). We first argue that conventional text-to-image (T2I) diffusion models, typically trained on paired image and text data (i.e., positive pairs) by minimizing score matching or flow matching losses, is suboptimal from the standpoint of representation alignment. Instead, a better alignment can be achieved through contrastive learning that leverages both positive and negative pairs. To achieve this efficiently even with pretrained models, we introduce a lightweight contrastive fine tuning strategy called SoftREPA that uses soft text tokens. This approach improves alignment with minimal computational overhead by adding fewer than 1M trainable parameters to the pretrained model. Our theoretical analysis demonstrates that our method explicitly increases the mutual information between text and image representations, leading to enhanced semantic consistency. Experimental results across text-to-image generation and text-guided image editing tasks validate the effectiveness of our approach in improving the semantic consistency of T2I generative models.
Authors: Yuanxin Liu, Rui Zhu, Shuhuai Ren, Jiacong Wang, Haoyuan Guo, Xu Sun, Lu Jiang
Abstract: With the rapid growth of video generative models (VGMs), it is essential to develop reliable and comprehensive automatic metrics for AI-generated videos (AIGVs). Existing methods either use off-the-shelf models optimized for other tasks or rely on human assessment data to train specialized evaluators. These approaches are constrained to specific evaluation aspects and are difficult to scale with the increasing demands for finer-grained and more comprehensive evaluations. To address this issue, this work investigates the feasibility of using multimodal large language models (MLLMs) as a unified evaluator for AIGVs, leveraging their strong visual perception and language understanding capabilities. To evaluate the performance of automatic metrics in unified AIGV evaluation, we introduce a benchmark called UVE-Bench. UVE-Bench collects videos generated by state-of-the-art VGMs and provides pairwise human preference annotations across 15 evaluation aspects. Using UVE-Bench, we extensively evaluate 16 MLLMs. Our empirical results suggest that while advanced MLLMs (e.g., Qwen2VL-72B and InternVL2.5-78B) still lag behind human evaluators, they demonstrate promising ability in unified AIGV evaluation, significantly surpassing existing specialized evaluation methods. Additionally, we conduct an in-depth analysis of key design choices that impact the performance of MLLM-driven evaluators, offering valuable insights for future research on AIGV evaluation. The code is available at https://github.com/bytedance/UVE.
Authors: Jialin Zhu, Jiangbei Yue, Feixiang He, He Wang
Abstract: Recently, 3D Gaussian Splatting (3DGS) provides a new framework for novel view synthesis, and has spiked a new wave of research in neural rendering and related applications. As 3DGS is becoming a foundational component of many models, any improvement on 3DGS itself can bring huge benefits. To this end, we aim to improve the fundamental paradigm and formulation of 3DGS. We argue that as an unnormalized mixture model, it needs to be neither Gaussians nor splatting. We subsequently propose a new mixture model consisting of flexible Student's t distributions, with both positive (splatting) and negative (scooping) densities. We name our model Student Splatting and Scooping, or SSS. When providing better expressivity, SSS also poses new challenges in learning. Therefore, we also propose a new principled sampling approach for optimization. Through exhaustive evaluation and comparison, across multiple datasets, settings, and metrics, we demonstrate that SSS outperforms existing methods in terms of quality and parameter efficiency, e.g. achieving matching or better quality with similar numbers of components, and obtaining comparable results while reducing the component number by as much as 82%.
Authors: Jinhao Duan, Fei Kong, Hao Cheng, James Diffenderfer, Bhavya Kailkhura, Lichao Sun, Xiaofeng Zhu, Xiaoshuang Shi, Kaidi Xu
Abstract: Object Hallucination (OH) has been acknowledged as one of the major trustworthy challenges in Large Vision-Language Models (LVLMs). Recent advancements in Large Language Models (LLMs) indicate that internal states, such as hidden states, encode the "overall truthfulness" of generated responses. However, it remains under-explored how internal states in LVLMs function and whether they could serve as "per-token" hallucination indicators, which is essential for mitigating OH. In this paper, we first conduct an in-depth exploration of LVLM internal states in relation to OH issues and discover that (1) LVLM internal states are high-specificity per-token indicators of hallucination behaviors. Moreover, (2) different LVLMs encode universal patterns of hallucinations in common latent subspaces, indicating that there exist "generic truthful directions" shared by various LVLMs. Based on these discoveries, we propose Truthful-Guided Pre-Intervention (TruthPrInt) that first learns the truthful direction of LVLM decoding and then applies truthful-guided inference-time intervention during LVLM decoding. We further propose ComnHallu to enhance both cross-LVLM and cross-data hallucination detection transferability by constructing and aligning hallucination latent subspaces. We evaluate TruthPrInt in extensive experimental settings, including in-domain and out-of-domain scenarios, over popular LVLMs and OH benchmarks. Experimental results indicate that TruthPrInt significantly outperforms state-of-the-art methods. Codes will be available at https://github.com/jinhaoduan/TruthPrInt.
Authors: Haoyuan Gao, Zicong Zhang, Yuqi Wei, Linglan Zhao, Guilin Li, Yexin Li, Linghe Kong, Weiran Huang
Abstract: Vision-Language Models (VLMs) represent a breakthrough in artificial intelligence by integrating visual and textual modalities to achieve impressive zero-shot capabilities. However, VLMs are susceptible to catastrophic forgetting when sequentially fine-tuned on multiple downstream tasks. Existing continual learning methods for VLMs often rely heavily on additional reference datasets, compromise zero-shot performance, or are limited to parameter-efficient fine-tuning scenarios. In this paper, we propose Continual Decoupling-Unifying (ConDU), a novel approach, by introducing model fusion into continual learning for VLMs. ConDU maintains a unified model along with task triggers and prototype sets, employing an iterative process of decoupling task-specific models for previous tasks and unifying them with the model for the newly learned task. Additionally, we introduce an inference strategy for zero-shot scenarios by aggregating predictions from multiple decoupled task-specific models. Extensive experiments across various settings show that ConDU achieves up to a 2\% improvement in average performance across all seen tasks compared to state-of-the-art baselines, while also enhancing zero-shot capabilities relative to the original VLM.
Authors: Jun Yu, Yang Zheng, Lei Wang, Yongqi Wang, Shengfan Xu
Abstract: Facial expression recognition is a challenging classification task that holds broad application prospects in the field of human-computer interaction. This paper aims to introduce the method we will adopt in the 8th Affective and Behavioral Analysis in the Wild (ABAW) Competition, which will be held during the Conference on Computer Vision and Pattern Recognition (CVPR) in 2025.First of all, we apply the frequency masking technique and the method of extracting data at equal time intervals to conduct targeted processing on the original videos. Then, based on the residual hybrid convolutional neural network and the multi-branch convolutional neural network respectively, we design feature extraction models for image and audio sequences. In particular, we propose a global channel-spatial attention mechanism to enhance the features initially extracted from both the audio and image modalities respectively.Finally, we adopt a decision fusion strategy based on the proportional criterion to fuse the classification results of the two single modalities, obtain an emotion probability vector, and output the final emotional classification. We also design a coarse - fine granularity loss function to optimize the performance of the entire network, which effectively improves the accuracy of facial expression recognition.In the facial expression recognition task of the 8th ABAW Competition, our method ranked third on the official validation set. This result fully confirms the effectiveness and competitiveness of the method we have proposed.
Authors: R. Gnana Praveen, Jahangir Alam, Eric Charton
Abstract: Audio and visual modalities are two predominant contact-free channels in videos, which are often expected to carry a complementary relationship with each other. However, they may not always complement each other, resulting in poor audio-visual feature representations. In this paper, we introduce Gated Recursive Joint Cross Attention (GRJCA) using a gating mechanism that can adaptively choose the most relevant features to effectively capture the synergic relationships across audio and visual modalities. Specifically, we improve the performance of Recursive Joint Cross-Attention (RJCA) by introducing a gating mechanism to control the flow of information between the input features and the attended features of multiple iterations depending on the strength of their complementary relationship. For instance, if the modalities exhibit strong complementary relationships, the gating mechanism emphasizes cross-attended features, otherwise non-attended features. To further improve the performance of the system, we also explored a hierarchical gating approach by introducing a gating mechanism at every iteration, followed by high-level gating across the gated outputs of each iteration. The proposed approach improves the performance of RJCA model by adding more flexibility to deal with weak complementary relationships across audio and visual modalities. Extensive experiments are conducted on the challenging Affwild2 dataset to demonstrate the robustness of the proposed approach. By effectively handling the weak complementary relationships across the audio and visual modalities, the proposed model achieves a Concordance Correlation Coefficient (CCC) of 0.561 (0.623) and 0.620 (0.660) for valence and arousal respectively on the test set (validation set).
Authors: Xiaoying Zhang, Da Peng, Yipeng Zhang, Zonghao Guo, Chengyue Wu, Chi Chen, Wei Ke, Helen Meng, Maosong Sun
Abstract: Despite their impressive capabilities, Multimodal Large Language Models (MLLMs) face challenges with fine-grained perception and complex reasoning. Prevalent multimodal pre-training approaches in MLLM construction focus on enhancing perception by training on high-quality image captions. While leveraging advanced MLLMs for caption generation enhances scalability, their outputs often lack comprehensiveness and accuracy. In this paper, we introduce Self-Improving cognition (SIcog), a self-learning framework designed to construct next-generation foundation MLLMs by enhancing their systematic cognitive capabilities through multimodal pre-training with self-generated data. Specifically, we propose Chain-of-Description (CoD), an approach that improves an MLLM's systematic perception by enabling step-by-step visual understanding. CoD sequentially focuses on salient content, fine-grained details, relational attributes, and peripheral context, before generating a coherent description, ensuring greater accuracy and comprehensiveness. Additionally, we adopt a structured chain-of-thought (CoT) reasoning technique to enable MLLMs to integrate in-depth multimodal reasoning. To construct a next-generation foundation MLLM with self-improved cognition, SIcog first equips an MLLM with systematic perception and reasoning abilities using minimal external annotations. The enhanced models then generate detailed captions and CoT reasoning data, which are further curated through self-consistency. This curated data is ultimately used for multimodal pre-training to develop next-generation foundation models. Extensive experiments on both low- and high-resolution MLLMs across diverse benchmarks demonstrate that, SIcog produces next-generation foundation MLLMs with significantly improved cognition, achieving benchmark-leading performance compared to prevalent pre-training approaches.
Authors: Kepeng Wu, Zecheng Li, Hezhen Hu, Wengang Zhou, Houqiang Li
Abstract: Pre-training has been proven to be effective in boosting the performance of Isolated Sign Language Recognition (ISLR). Existing pre-training methods solely focus on the compact pose data, which eliminates background perturbation but inevitably suffers from insufficient semantic cues compared to raw RGB videos. Nevertheless, learning representation directly from RGB videos remains challenging due to the presence of sign-independent visual features. To address this dilemma, we propose a Cross-modal Consistency Learning framework (CCL-SLR), which leverages the cross-modal consistency from both RGB and pose modalities based on self-supervised pre-training. First, CCL-SLR employs contrastive learning for instance discrimination within and across modalities. Through the single-modal and cross-modal contrastive learning, CCL-SLR gradually aligns the feature spaces of RGB and pose modalities, thereby extracting consistent sign representations. Second, we further introduce Motion-Preserving Masking (MPM) and Semantic Positive Mining (SPM) techniques to improve cross-modal consistency from the perspective of data augmentation and sample similarity, respectively. Extensive experiments on four ISLR benchmarks show that CCL-SLR achieves impressive performance, demonstrating its effectiveness. The code will be released to the public.
Authors: Jie Huang, Haorui Chen, Jiaxuan Ren, Siran Peng, Liangjian Deng
Abstract: Currently, deep learning-based methods for remote sensing pansharpening have advanced rapidly. However, many existing methods struggle to fully leverage feature heterogeneity and redundancy, thereby limiting their effectiveness. We use the covariance matrix to model the feature heterogeneity and redundancy and propose Correlation-Aware Covariance Weighting (CACW) to adjust them. CACW captures these correlations through the covariance matrix, which is then processed by a nonlinear function to generate weights for adjustment. Building upon CACW, we introduce a general adaptive dual-level weighting mechanism (ADWM) to address these challenges from two key perspectives, enhancing a wide range of existing deep-learning methods. First, Intra-Feature Weighting (IFW) evaluates correlations among channels within each feature to reduce redundancy and enhance unique information. Second, Cross-Feature Weighting (CFW) adjusts contributions across layers based on inter-layer correlations, refining the final output. Extensive experiments demonstrate the superior performance of ADWM compared to recent state-of-the-art (SOTA) methods. Furthermore, we validate the effectiveness of our approach through generality experiments, redundancy visualization, comparison experiments, key variables and complexity analysis, and ablation studies. Our code is available at https://github.com/Jie-1203/ADWM.
Authors: Eitan Shaar, Ariel Shaulov, Gal Chechik, Lior Wolf
Abstract: In the domain of audio-visual event perception, which focuses on the temporal localization and classification of events across distinct modalities (audio and visual), existing approaches are constrained by the vocabulary available in their training data. This limitation significantly impedes their capacity to generalize to novel, unseen event categories. Furthermore, the annotation process for this task is labor-intensive, requiring extensive manual labeling across modalities and temporal segments, limiting the scalability of current methods. Current state-of-the-art models ignore the shifts in event distributions over time, reducing their ability to adjust to changing video dynamics. Additionally, previous methods rely on late fusion to combine audio and visual information. While straightforward, this approach results in a significant loss of multimodal interactions. To address these challenges, we propose Audio-Visual Adaptive Video Analysis ($\text{AV}^2\text{A}$), a model-agnostic approach that requires no further training and integrates a score-level fusion technique to retain richer multimodal interactions. $\text{AV}^2\text{A}$ also includes a within-video label shift algorithm, leveraging input video data and predictions from prior frames to dynamically adjust event distributions for subsequent frames. Moreover, we present the first training-free, open-vocabulary baseline for audio-visual event perception, demonstrating that $\text{AV}^2\text{A}$ achieves substantial improvements over naive training-free baselines. We demonstrate the effectiveness of $\text{AV}^2\text{A}$ on both zero-shot and weakly-supervised state-of-the-art methods, achieving notable improvements in performance metrics over existing approaches.
Authors: Qingyao Xu, Siheng Chen, Guang Chen, Yanfeng Wang, Ya Zhang
Abstract: Traffic scene understanding is essential for intelligent transportation systems and autonomous driving, ensuring safe and efficient vehicle operation. While recent advancements in VLMs have shown promise for holistic scene understanding, the application of VLMs to traffic scenarios, particularly using BEV maps, remains under explored. Existing methods often suffer from limited task design and narrow data amount, hindering comprehensive scene understanding. To address these challenges, we introduce ChatBEV-QA, a novel BEV VQA benchmark contains over 137k questions, designed to encompass a wide range of scene understanding tasks, including global scene understanding, vehicle-lane interactions, and vehicle-vehicle interactions. This benchmark is constructed using an novel data collection pipeline that generates scalable and informative VQA data for BEV maps. We further fine-tune a specialized vision-language model ChatBEV, enabling it to interpret diverse question prompts and extract relevant context-aware information from BEV maps. Additionally, we propose a language-driven traffic scene generation pipeline, where ChatBEV facilitates map understanding and text-aligned navigation guidance, significantly enhancing the generation of realistic and consistent traffic scenarios. The dataset, code and the fine-tuned model will be released.
Authors: Jiang Qin, Senmao Li, Alexandra Gomez-Villa, Shiqi Yang, Yaxing Wang, Kai Wang, Joost van de Weijer
Abstract: Recent advances in Text-to-Image (T2I) diffusion models have transformed image generation, enabling significant progress in stylized generation using only a few style reference images. However, current diffusion-based methods struggle with fine-grained style customization due to challenges in controlling multiple style attributes, such as color and texture. This paper introduces the first tuning-free approach to achieve free-lunch color-texture disentanglement in stylized T2I generation, addressing the need for independently controlled style elements for the Disentangled Stylized Image Generation (DisIG) problem. Our approach leverages the Image-Prompt Additivity property in the CLIP image embedding space to develop techniques for separating and extracting Color-Texture Embeddings (CTE) from individual color and texture reference images. To ensure that the color palette of the generated image aligns closely with the color reference, we apply a whitening and coloring transformation to enhance color consistency. Additionally, to prevent texture loss due to the signal-leak bias inherent in diffusion training, we introduce a noise term that preserves textural fidelity during the Regularized Whitening and Coloring Transformation (RegWCT). Through these methods, our Style Attributes Disentanglement approach (SADis) delivers a more precise and customizable solution for stylized image generation. Experiments on images from the WikiArt and StyleDrop datasets demonstrate that, both qualitatively and quantitatively, SADis surpasses state-of-the-art stylization methods in the DisIG task.Code will be released at https://deepffff.github.io/sadis.github.io/.
Authors: Yi Du, Zhipeng Zhao, Shaoshu Su, Sharath Golluri, Haoze Zheng, Runmao Yao, Chen Wang
Abstract: Point cloud (PC) processing tasks-such as completion, upsampling, denoising, and colorization-are crucial in applications like autonomous driving and 3D reconstruction. Despite substantial advancements, prior approaches often address each of these tasks independently, with separate models focused on individual issues. However, this isolated approach fails to account for the fact that defects like incompleteness, low resolution, noise, and lack of color frequently coexist, with each defect influencing and correlating with the others. Simply applying these models sequentially can lead to error accumulation from each model, along with increased computational costs. To address these challenges, we introduce SuperPC, the first unified diffusion model capable of concurrently handling all four tasks. Our approach employs a three-level-conditioned diffusion framework, enhanced by a novel spatial-mix-fusion strategy, to leverage the correlations among these four defects for simultaneous, efficient processing. We show that SuperPC outperforms the state-of-the-art specialized models as well as their combination on all four individual tasks.
Authors: Parker Ewen, Hao Chen, Seth Isaacson, Joey Wilson, Katherine A. Skinner, Ram Vasudevan
Abstract: This paper introduces a novel approach to uncertainty quantification for radiance fields by leveraging higher-order moments of the rendering equation. Uncertainty quantification is crucial for downstream tasks including view planning and scene understanding, where safety and robustness are paramount. However, the high dimensionality and complexity of radiance fields pose significant challenges for uncertainty quantification, limiting the use of these uncertainty quantification methods in high-speed decision-making. We demonstrate that the probabilistic nature of the rendering process enables efficient and differentiable computation of higher-order moments for radiance field outputs, including color, depth, and semantic predictions. Our method outperforms existing radiance field uncertainty estimation techniques while offering a more direct, computationally efficient, and differentiable formulation without the need for post-processing. Beyond uncertainty quantification, we also illustrate the utility of our approach in downstream applications such as next-best-view (NBV) selection and active ray sampling for neural radiance field training. Extensive experiments on synthetic and real-world scenes confirm the efficacy of our approach, which achieves state-of-the-art performance while maintaining simplicity.
Authors: Vaibhav Rathore, Shubhranil B, Saikat Dutta, Sarthak Mehrotra, Zsolt Kira, Biplab Banerjee
Abstract: Generalized Class Discovery (GCD) clusters base and novel classes in a target domain using supervision from a source domain with only base classes. Current methods often falter with distribution shifts and typically require access to target data during training, which can sometimes be impractical. To address this issue, we introduce the novel paradigm of Domain Generalization in GCD (DG-GCD), where only source data is available for training, while the target domain, with a distinct data distribution, remains unseen until inference. To this end, our solution, DG2CD-Net, aims to construct a domain-independent, discriminative embedding space for GCD. The core innovation is an episodic training strategy that enhances cross-domain generalization by adapting a base model on tasks derived from source and synthetic domains generated by a foundation model. Each episode focuses on a cross-domain GCD task, diversifying task setups over episodes and combining open-set domain adaptation with a novel margin loss and representation learning for optimizing the feature space progressively. To capture the effects of fine-tuning on the base model, we extend task arithmetic by adaptively weighting the local task vectors concerning the fine-tuned models based on their GCD performance on a validation distribution. This episodic update mechanism boosts the adaptability of the base model to unseen targets. Experiments across three datasets confirm that DG2CD-Net outperforms existing GCD methods customized for DG-GCD.
Authors: Weixiao Gao, Liangliang Nan, Hugo Ledoux
Abstract: Semantic segmentation in urban scene analysis has mainly focused on images or point clouds, while textured meshes - offering richer spatial representation - remain underexplored. This paper introduces SUM Parts, the first large-scale dataset for urban textured meshes with part-level semantic labels, covering about 2.5 km2 with 21 classes. The dataset was created using our own annotation tool, which supports both face- and texture-based annotations with efficient interactive selection. We also provide a comprehensive evaluation of 3D semantic segmentation and interactive annotation methods on this dataset. Our project page is available at https://tudelft3d.github.io/SUMParts/.
Authors: Debabrata Mandal, Soumitri Chattopadhyay, Guansen Tong, Praneeth Chakravarthula
Abstract: Image restoration is essential for enhancing degraded images across computer vision tasks. However, most existing methods address only a single type of degradation (e.g., blur, noise, or haze) at a time, limiting their real-world applicability where multiple degradations often occur simultaneously. In this paper, we propose UniCoRN, a unified image restoration approach capable of handling multiple degradation types simultaneously using a multi-head diffusion model. Specifically, we uncover the potential of low-level visual cues extracted from images in guiding a controllable diffusion model for real-world image restoration and we design a multi-head control network adaptable via a mixture-of-experts strategy. We train our model without any prior assumption of specific degradations, through a smartly designed curriculum learning recipe. Additionally, we also introduce MetaRestore, a metalens imaging benchmark containing images with multiple degradations and artifacts. Extensive evaluations on several challenging datasets, including our benchmark, demonstrate that our method achieves significant performance gains and can robustly restore images with severe degradations. Project page: https://codejaeger.github.io/unicorn-gh
Authors: Hui Liu, Wenya Wang, Kecheng Chen, Jie Liu, Yibing Liu, Tiexin Qin, Peisong He, Xinghao Jiang, Haoliang Li
Abstract: In zero-shot image recognition tasks, humans demonstrate remarkable flexibility in classifying unseen categories by composing known simpler concepts. However, existing vision-language models (VLMs), despite achieving significant progress through large-scale natural language supervision, often underperform in real-world applications because of sub-optimal prompt engineering and the inability to adapt effectively to target classes. To address these issues, we propose a Concept-guided Human-like Bayesian Reasoning (CHBR) framework. Grounded in Bayes' theorem, CHBR models the concept used in human image recognition as latent variables and formulates this task by summing across potential concepts, weighted by a prior distribution and a likelihood function. To tackle the intractable computation over an infinite concept space, we introduce an importance sampling algorithm that iteratively prompts large language models (LLMs) to generate discriminative concepts, emphasizing inter-class differences. We further propose three heuristic approaches involving Average Likelihood, Confidence Likelihood, and Test Time Augmentation (TTA) Likelihood, which dynamically refine the combination of concepts based on the test image. Extensive evaluations across fifteen datasets demonstrate that CHBR consistently outperforms existing state-of-the-art zero-shot generalization methods.
Authors: Chun-Han Yao, Yiming Xie, Vikram Voleti, Huaizu Jiang, Varun Jampani
Abstract: We present Stable Video 4D 2.0 (SV4D 2.0), a multi-view video diffusion model for dynamic 3D asset generation. Compared to its predecessor SV4D, SV4D 2.0 is more robust to occlusions and large motion, generalizes better to real-world videos, and produces higher-quality outputs in terms of detail sharpness and spatio-temporal consistency. We achieve this by introducing key improvements in multiple aspects: 1) network architecture: eliminating the dependency of reference multi-views and designing blending mechanism for 3D and frame attention, 2) data: enhancing quality and quantity of training data, 3) training strategy: adopting progressive 3D-4D training for better generalization, and 4) 4D optimization: handling 3D inconsistency and large motion via 2-stage refinement and progressive frame sampling. Extensive experiments demonstrate significant performance gain by SV4D 2.0 both visually and quantitatively, achieving better detail (-14\% LPIPS) and 4D consistency (-44\% FV4D) in novel-view video synthesis and 4D optimization (-12\% LPIPS and -24\% FV4D) compared to SV4D.
Authors: Zahra Shamsi, Isaac Reid, Drew Bryant, Jacob Wilson, Xiaoyu Qu, Avinava Dubey, Konik Kothari, Mostafa Dehghani, Mariya Chavarha, Valerii Likhosherstov, Brian Williams, Michael Frumkin, Fred Appelbaum, Krzysztof Choromanski, Ali Bashir, Min Fang
Abstract: We present a machine learning method capable of accurately detecting chromosome abnormalities that cause blood cancers directly from microscope images of the metaphase stage of cell division. The pipeline is built on a series of fine-tuned Vision Transformers. Current state of the art (and standard clinical practice) requires expensive, manual expert analysis, whereas our pipeline takes only 15 seconds per metaphase image. Using a novel pretraining-finetuning strategy to mitigate the challenge of data scarcity, we achieve a high precision-recall score of 94% AUC for the clinically significant del(5q) and t(9;22) anomalies. Our method also unlocks zero-shot detection of rare aberrations based on model latent embeddings. The ability to quickly, accurately, and scalably diagnose genetic abnormalities directly from metaphase images could transform karyotyping practice and improve patient outcomes. We will make code publicly available.
Authors: Andrew Yanzhe Ke, Lei Luo, Alexandre Chapiro, Xiaoyu Xiang, Yuchen Fan, Rakesh Ranjan, Rafal Mantiuk
Abstract: The vast majority of standard image and video content available online is represented in display-encoded color spaces, in which pixel values are conveniently scaled to a limited range (0-1) and the color distribution is approximately perceptually uniform. In contrast, both camera RAW and high dynamic range (HDR) images are often represented in linear color spaces, in which color values are linearly related to colorimetric quantities of light. While training on commonly available display-encoded images is a well-established practice, there is no consensus on how neural networks should be trained for tasks on RAW and HDR images in linear color spaces. In this work, we test several approaches on three popular image restoration applications: denoising, deblurring, and single-image super-resolution. We examine whether HDR/RAW images need to be display-encoded using popular transfer functions (PQ, PU21, and mu-law), or whether it is better to train in linear color spaces, but use loss functions that correct for perceptual non-uniformity. Our results indicate that neural networks train significantly better on HDR and RAW images represented in display-encoded color spaces, which offer better perceptual uniformity than linear spaces. This small change to the training strategy can bring a very substantial gain in performance, between 2 and 9 dB.
Authors: Yinuo Zhao, Kun Wu, Tianjiao Yi, Zhiyuan Xu, Xiaozhu Ju, Zhengping Che, Chi Harold Liu, Jian Tang
Abstract: Improving generalization is one key challenge in embodied AI, where obtaining large-scale datasets across diverse scenarios is costly. Traditional weak augmentations, such as cropping and flipping, are insufficient for improving a model's performance in new environments. Existing data augmentation methods often disrupt task-relevant information in images, potentially degrading performance. To overcome these challenges, we introduce EAGLE, an efficient training framework for generalizable visuomotor policies that improves upon existing methods by (1) enhancing generalization by applying augmentation only to control-related regions identified through a self-supervised control-aware mask and (2) improving training stability and efficiency by distilling knowledge from an expert to a visuomotor student policy, which is then deployed to unseen environments without further fine-tuning. Comprehensive experiments on three domains, including the DMControl Generalization Benchmark, the enhanced Robot Manipulation Distraction Benchmark, and a long-sequential drawer-opening task, validate the effectiveness of our method.
Authors: George Yiasemis, Jan-Jakob Sonke, Jonas Teuwen
Abstract: $\textbf{Background:}$ Accelerating dynamic MRI is vital for advancing clinical applications and improving patient comfort. Commonly, deep learning (DL) methods for accelerated dynamic MRI reconstruction typically rely on uniformly applying non-adaptive predetermined or random subsampling patterns across all temporal frames of the dynamic acquisition. This approach fails to exploit temporal correlations or optimize subsampling on a case-by-case basis. $\textbf{Purpose:}$ To develop an end-to-end approach for adaptive dynamic MRI subsampling and reconstruction, capable of generating customized sampling patterns maximizing at the same time reconstruction quality. $\textbf{Methods:}$ We introduce the End-to-end Adaptive Dynamic Sampling and Reconstruction (E2E-ADS-Recon) for MRI framework, which integrates an adaptive dynamic sampler (ADS) that adapts the acquisition trajectory to each case for a given acceleration factor with a state-of-the-art dynamic reconstruction network, vSHARP, for reconstructing the adaptively sampled data into a dynamic image. The ADS can produce either frame-specific patterns or unified patterns applied to all temporal frames. E2E-ADS-Recon is evaluated under both frame-specific and unified 1D or 2D sampling settings, using dynamic cine cardiac MRI data and compared with vSHARP models employing standard subsampling trajectories, as well as pipelines where ADS was replaced by parameterized samplers optimized for dataset-specific schemes. $\textbf{Results:}$ E2E-ADS-Recon exhibited superior reconstruction quality, especially at high accelerations, in terms of standard quantitative metrics (SSIM, pSNR, NMSE). $\textbf{Conclusion:}$ The proposed framework improves reconstruction quality, highlighting the importance of case-specific subsampling optimization in dynamic MRI applications.
Authors: Shenghong Dai, Shiqi Jiang, Yifan Yang, Ting Cao, Mo Li, Suman Banerjee, Lili Qiu
Abstract: This paper presents Babel, the expandable modality alignment model, specially designed for multi-modal sensing. While there has been considerable work on multi-modality alignment, they all struggle to effectively incorporate multiple sensing modalities due to the data scarcity constraints. How to utilize multi-modal data with partial pairings in sensing remains an unresolved challenge. Babel tackles this challenge by introducing the concept of expandable modality alignment. The key idea involves transforming the N-modality alignment into a series of binary-modality alignments. Novel techniques are also proposed to further mitigate data scarcity issue and balance the contribution of the newly incorporated modality with the previously established modality alignment during the expandable alignment process. We provide the comprehensive implementation. In the pre-training phase, Babel currently aligns 6 sensing modalities, namely Wi-Fi, mmWave, IMU, LiDAR, video, and depth. For the deployment phase, as a foundation model, any single or combination of aligned modalities could be selected from Babel and applied to downstream tasks. Evaluation demonstrates Babel's outstanding performance on eight human activity recognition datasets, compared to a broad range of baselines e.g., the SOTA single-modal sensing networks, multi-modal sensing framework, and multi-modal large language models. Babel not only improves the performance of individual modality sensing (12% averaged accuracy improvement), but also effectively fuses multiple available modalities (up to 22% accuracy increase). Case studies also highlight emerging application scenarios empowered by Babel, including cross-modality retrieval (i.e., sensing imaging), and bridging LLM for sensing comprehension.
Authors: Junchao Zhu, Mengmeng Yin, Ruining Deng, Yitian Long, Yu Wang, Yaohong Wang, Shilin Zhao, Haichun Yang, Yuankai Huo
Abstract: Accurate delineation of the boundaries between the renal cortex and medulla is crucial for subsequent functional structural analysis and disease diagnosis. Training high-quality deep-learning models for layer segmentation relies on the availability of large amounts of annotated data. However, due to the patient's privacy of medical data and scarce clinical cases, constructing pathological datasets from clinical sources is relatively difficult and expensive. Moreover, using external natural image datasets introduces noise during the domain generalization process. Cross-species homologous data, such as mouse kidney data, which exhibits high structural and feature similarity to human kidneys, has the potential to enhance model performance on human datasets. In this study, we incorporated the collected private Periodic Acid-Schiff (PAS) stained mouse kidney dataset into the human kidney dataset for joint training. The results showed that after introducing cross-species homologous data, the semantic segmentation models based on CNN and Transformer architectures achieved an average increase of 1.77% and 1.24% in mIoU, and 1.76% and 0.89% in Dice score for the human renal cortex and medulla datasets, respectively. This approach is also capable of enhancing the model's generalization ability. This indicates that cross-species homologous data, as a low-noise trainable data source, can help improve model performance under conditions of limited clinical samples. Code is available at https://github.com/hrlblab/layer_segmentation.
Authors: Zongzhe Xu, Ritvik Gupta, Wenduo Cheng, Alexander Shen, Junhong Shen, Ameet Talwalkar, Mikhail Khodak
Abstract: Following its success for vision and text, the "foundation model" (FM) paradigm -- pretraining large models on massive data, then fine-tuning on target tasks -- has rapidly expanded to domains in the sciences, engineering, healthcare, and beyond. Has this achieved what the original FMs accomplished, i.e. the supplanting of traditional supervised learning in their domains? To answer we look at three modalities -- genomics, satellite imaging, and time series -- with multiple recent FMs and compare them to a standard supervised learning workflow: model development, hyperparameter tuning, and training, all using only data from the target task. Across these three specialized domains, we find that it is consistently possible to train simple supervised models -- no more complicated than a lightly modified wide ResNet or UNet -- that match or even outperform the latest foundation models. Our work demonstrates that the benefits of large-scale pretraining have yet to be realized in many specialized areas, reinforces the need to compare new FMs to strong, well-tuned baselines, and introduces two new, easy-to-use, open-source, and automated workflows for doing so.
Authors: Owais Ahmed, M Huzaifa, M Areeb, Hamza Ali Khan
Abstract: In response to the growing challenges of manual labor and efficiency in warehouse operations, Amazon has embarked on a significant transformation by incorporating robotics to assist with various tasks. While a substantial number of robots have been successfully deployed for tasks such as item transportation within warehouses, the complex process of object picking from shelves remains a significant challenge. This project addresses the issue by developing an innovative robotic system capable of autonomously fulfilling a simulated order by efficiently selecting specific items from shelves. A distinguishing feature of the proposed robotic system is its capacity to navigate the challenge of uncertain object positions within each bin of the shelf. The system is engineered to autonomously adapt its approach, employing strategies that enable it to efficiently locate and retrieve the desired items, even in the absence of pre-established knowledge about their placements.
Authors: George Yiasemis, Jan-Jakob Sonke, Jonas Teuwen
Abstract: Dynamic MRI enables a range of clinical applications, including cardiac function assessment, organ motion tracking, and radiotherapy guidance. However, fully sampling the dynamic k-space data is often infeasible due to time constraints and physiological motion such as respiratory and cardiac motion. This necessitates undersampling, which degrades the quality of reconstructed images. Poor image quality not only hinders visualization but also impairs the estimation of deformation fields, crucial for registering dynamic (moving) images to a static reference image. This registration enables tasks such as motion correction, treatment planning, and quantitative analysis in applications like cardiac imaging and MR-guided radiotherapy. To overcome the challenges posed by undersampling and motion, we introduce an end-to-end deep learning (DL) framework that integrates adaptive dynamic k-space sampling, reconstruction, and registration. Our approach begins with a DL-based adaptive sampling strategy, optimizing dynamic k-space acquisition to capture the most relevant data for each specific case. This is followed by a DL-based reconstruction module that produces images optimized for accurate deformation field estimation from the undersampled moving data. Finally, a registration module estimates the deformation fields aligning the reconstructed dynamic images with a static reference. The proposed framework is independent of specific reconstruction and registration modules allowing for plug-and-play integration of these components. The entire framework is jointly trained using a combination of supervised and unsupervised loss functions, enabling end-to-end optimization for improved performance across all components. Through controlled experiments and ablation studies, we validate each component, demonstrating that each choice contributes to robust motion estimation from undersampled dynamic data.
Authors: Shufan Li, Konstantinos Kallidromitis, Akash Gokul, Zichun Liao, Yusuke Kato, Kazuki Kozuka, Aditya Grover
Abstract: We introduce OmniFlow, a novel generative model designed for any-to-any generation tasks such as text-to-image, text-to-audio, and audio-to-image synthesis. OmniFlow advances the rectified flow (RF) framework used in text-to-image models to handle the joint distribution of multiple modalities. It outperforms previous any-to-any models on a wide range of tasks, such as text-to-image and text-to-audio synthesis. Our work offers three key contributions: First, we extend RF to a multi-modal setting and introduce a novel guidance mechanism, enabling users to flexibly control the alignment between different modalities in the generated outputs. Second, we propose a novel architecture that extends the text-to-image MMDiT architecture of Stable Diffusion 3 and enables audio and text generation. The extended modules can be efficiently pretrained individually and merged with the vanilla text-to-image MMDiT for fine-tuning. Lastly, we conduct a comprehensive study on the design choices of rectified flow transformers for large-scale audio and text generation, providing valuable insights into optimizing performance across diverse modalities. The Code will be available at https://github.com/jacklishufan/OmniFlows.
Authors: Yi Chen, Yuying Ge, Weiliang Tang, Yizhuo Li, Yixiao Ge, Mingyu Ding, Ying Shan, Xihui Liu
Abstract: Recent developments in Large Language Models pre-trained on extensive corpora have shown significant success in various natural language processing tasks with minimal fine-tuning. This success offers new promise for robotics, which has long been constrained by the high cost of action-labeled data. We ask: given the abundant video data containing interaction-related knowledge available as a rich "corpus", can a similar generative pre-training approach be effectively applied to enhance robot learning? The key challenge is to identify an effective representation for autoregressive pre-training that benefits robot manipulation tasks. Inspired by the way humans learn new skills through observing dynamic environments, we propose that effective robotic learning should emphasize motion-related knowledge, which is closely tied to low-level actions and is hardware-agnostic, facilitating the transfer of learned motions to actual robot actions. To this end, we introduce Moto, which converts video content into latent Motion Token sequences by a Latent Motion Tokenizer, learning a bridging "language" of motion from videos in an unsupervised manner. We pre-train Moto-GPT through motion token autoregression, enabling it to capture diverse visual motion knowledge. After pre-training, Moto-GPT demonstrates the promising ability to produce semantically interpretable motion tokens, predict plausible motion trajectories, and assess trajectory rationality through output likelihood. To transfer learned motion priors to real robot actions, we implement a co-fine-tuning strategy that seamlessly bridges latent motion token prediction and real robot control. Extensive experiments show that the fine-tuned Moto-GPT exhibits superior robustness and efficiency on robot manipulation benchmarks, underscoring its effectiveness in transferring knowledge from video data to downstream visual manipulation tasks.
Authors: Enshen Zhou, Qi Su, Cheng Chi, Zhizheng Zhang, Zhongyuan Wang, Tiejun Huang, Lu Sheng, He Wang
Abstract: Automatic detection and prevention of open-set failures are crucial in closed-loop robotic systems. Recent studies often struggle to simultaneously identify unexpected failures reactively after they occur and prevent foreseeable ones proactively. To this end, we propose Code-as-Monitor (CaM), a novel paradigm leveraging the vision-language model (VLM) for both open-set reactive and proactive failure detection. The core of our method is to formulate both tasks as a unified set of spatio-temporal constraint satisfaction problems and use VLM-generated code to evaluate them for real-time monitoring. To enhance the accuracy and efficiency of monitoring, we further introduce constraint elements that abstract constraint-related entities or their parts into compact geometric elements. This approach offers greater generality, simplifies tracking, and facilitates constraint-aware visual programming by leveraging these elements as visual prompts. Experiments show that CaM achieves a 28.7% higher success rate and reduces execution time by 31.8% under severe disturbances compared to baselines across three simulators and a real-world setting. Moreover, CaM can be integrated with open-loop control policies to form closed-loop systems, enabling long-horizon tasks in cluttered scenes with dynamic environments.
Authors: Simran Khanuja, Vivek Iyer, Claire He, Graham Neubig
Abstract: Beyond conventional paradigms of translating speech and text, recently, there has been interest in automated transcreation of images to facilitate localization of visual content across different cultures. Attempts to define this as a formal Machine Learning (ML) problem have been impeded by the lack of automatic evaluation mechanisms, with previous work relying solely on human evaluation. In this paper, we seek to close this gap by proposing a suite of automatic evaluation metrics inspired by machine translation (MT) metrics, categorized into: a) Object-based, b) Embedding-based, and c) VLM-based. Drawing on theories from translation studies and real-world transcreation practices, we identify three critical dimensions of image transcreation: cultural relevance, semantic equivalence and visual similarity, and design our metrics to evaluate systems along these axes. Our results show that proprietary VLMs best identify cultural relevance and semantic equivalence, while vision-encoder representations are adept at measuring visual similarity. Meta-evaluation across 7 countries shows our metrics agree strongly with human ratings, with average segment-level correlations ranging from 0.55-0.87. Finally, through a discussion of the merits and demerits of each metric, we offer a robust framework for automated image transcreation evaluation, grounded in both theoretical foundations and practical application. Our code can be found here: https://github.com/simran-khanuja/automatic-eval-img-transcreation.
URLs: https://github.com/simran-khanuja/automatic-eval-img-transcreation.
Authors: JunEn Low, Maximilian Adang, Javier Yu, Keiko Nagami, Mac Schwager
Abstract: We propose a new simulator, training approach, and policy architecture, collectively called SOUS VIDE, for end-to-end visual drone navigation. Our trained policies exhibit zero-shot sim-to-real transfer with robust real-world performance using only onboard perception and computation. Our simulator, called FiGS, couples a computationally simple drone dynamics model with a high visual fidelity Gaussian Splatting scene reconstruction. FiGS can quickly simulate drone flights producing photorealistic images at up to 130 fps. We use FiGS to collect 100k-300k image/state-action pairs from an expert MPC with privileged state and dynamics information, randomized over dynamics parameters and spatial disturbances. We then distill this expert MPC into an end-to-end visuomotor policy with a lightweight neural architecture, called SV-Net. SV-Net processes color image, optical flow and IMU data streams into low-level thrust and body rate commands at 20 Hz onboard a drone. Crucially, SV-Net includes a learned module for low-level control that adapts at runtime to variations in drone dynamics. In a campaign of 105 hardware experiments, we show SOUS VIDE policies to be robust to 30% mass variations, 40 m/s wind gusts, 60% changes in ambient brightness, shifting or removing objects from the scene, and people moving aggressively through the drone's visual field. Code, data, and experiment videos can be found on our project page: https://stanfordmsl.github.io/SousVide/.
Authors: Jiarui Wu, Zhuo Liu, Hangfeng He
Abstract: Spatial relation hallucinations pose a persistent challenge in large vision-language models (LVLMs), leading to generate incorrect predictions about object positions and spatial configurations within an image. To address this issue, we propose a constraint-aware prompting framework designed to reduce spatial relation hallucinations. Specifically, we introduce two types of constraints: (1) bidirectional constraint, which ensures consistency in pairwise object relations, and (2) transitivity constraint, which enforces relational dependence across multiple objects. By incorporating these constraints, LVLMs can produce more spatially coherent and consistent outputs. We evaluate our method on three widely-used spatial relation datasets, demonstrating performance improvements over existing approaches. Additionally, a systematic analysis of various bidirectional relation analysis choices and transitivity reference selections highlights greater possibilities of our methods in incorporating constraints to mitigate spatial relation hallucinations.
Authors: Jiantao Lin, Xin Yang, Meixi Chen, Yingjie Xu, Dongyu Yan, Leyi Wu, Xinli Xu, Lie XU, Shunsi Zhang, Ying-Cong Chen
Abstract: Diffusion models have achieved great success in generating 2D images. However, the quality and generalizability of 3D content generation remain limited. State-of-the-art methods often require large-scale 3D assets for training, which are challenging to collect. In this work, we introduce Kiss3DGen (Keep It Simple and Straightforward in 3D Generation), an efficient framework for generating, editing, and enhancing 3D objects by repurposing a well-trained 2D image diffusion model for 3D generation. Specifically, we fine-tune a diffusion model to generate ''3D Bundle Image'', a tiled representation composed of multi-view images and their corresponding normal maps. The normal maps are then used to reconstruct a 3D mesh, and the multi-view images provide texture mapping, resulting in a complete 3D model. This simple method effectively transforms the 3D generation problem into a 2D image generation task, maximizing the utilization of knowledge in pretrained diffusion models. Furthermore, we demonstrate that our Kiss3DGen model is compatible with various diffusion model techniques, enabling advanced features such as 3D editing, mesh and texture enhancement, etc. Through extensive experiments, we demonstrate the effectiveness of our approach, showcasing its ability to produce high-quality 3D models efficiently.
Authors: Alvin Kimbowa, Arjun Parmar, Maziar Badii, David Liu, Matthew Harkey, Ilker Hacihaliloglu
Abstract: Automated knee cartilage segmentation using point-of-care ultrasound devices and deep-learning networks has the potential to enhance the management of knee osteoarthritis. However, segmentation algorithms often struggle with domain shifts caused by variations in ultrasound devices and acquisition parameters, limiting their generalizability. In this paper, we propose Mono2D, a monogenic layer that extracts multi-scale, contrast- and intensity-invariant local phase features using trainable bandpass quadrature filters. This layer mitigates domain shifts, improving generalization to out-of-distribution domains. Mono2D is integrated before the first layer of a segmentation network, and its parameters jointly trained alongside the network's parameters. We evaluated Mono2D on a multi-domain 2D ultrasound knee cartilage dataset for single-source domain generalization (SSDG). Our results demonstrate that Mono2D outperforms other SSDG methods in terms of Dice score and mean average surface distance. To further assess its generalizability, we evaluate Mono2D on a multi-site prostate MRI dataset, where it continues to outperform other SSDG methods, highlighting its potential to improve domain generalization in medical imaging. Nevertheless, further evaluation on diverse datasets is still necessary to assess its clinical utility.