Computer Vision
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
Tuesday 25 November
Weight Averaging (WA) has emerged as a powerful technique for enhancing generalization by promoting convergence to a flat loss landscape, which correlates with stronger out-of-distribution performance. However, applying WA directly to multi-modal domain generalization (MMDG) is challenging: differences in optimization speed across modalities lead WA to overfit to faster-converging ones in early stages, suppressing the contribution of slower yet complementary modalities, thereby hindering effective modality fusion and skewing the loss surface toward sharper, less generalizable minima. To address this issue, we propose MBCD, a unified collaborative distillation framework that retains WA's flatness-inducing advantages while overcoming its shortcomings in multi-modal contexts. MBCD begins with adaptive modality dropout in the student model to curb early-stage bias toward dominant modalities. A gradient consistency constraint then aligns learning signals between uni-modal branches and the fused representation, encouraging coordinated and smoother optimization. Finally, a WA-based teacher conducts cross-modal distillation by transferring fused knowledge to each uni-modal branch, which strengthens cross-modal interactions and steer convergence toward flatter solutions. Extensive experiments on MMDG benchmarks show that MBCD consistently outperforms existing methods, achieving superior accuracy and robustness across diverse unseen domains.
Computer Vision
The Image as Its Own Reward: Reinforcement Learning with Adversarial Reward for Image Generation
A reliable reward function is essential for reinforcement learning (RL) in image generation. Most current RL approaches depend on pre-trained preference models that output scalar rewards to approximate human preferences. However, these rewards often fail to capture human perception and are vulnerable to reward hacking, where higher scores do not correspond to better images. To address this, we introduce Adv-GRPO, an RL framework with an adversarial reward that iteratively updates both the reward model and the generator. The reward model is supervised using reference images as positive samples and can largely avoid being hacked. Unlike KL regularization that constrains parameter updates, our learned reward directly guides the generator through its visual outputs, leading to higher-quality images. Moreover, while optimizing existing reward functions can alleviate reward hacking, their inherent biases remain. For instance, PickScore may degrade image quality, whereas OCR-based rewards often reduce aesthetic fidelity. To address this, we take the image itself as a reward, using reference images and vision foundation models (e.g., DINO) to provide rich visual rewards. These dense visual signals, instead of a single scalar, lead to consistent gains across image quality, aesthetics, and task-specific metrics. Finally, we show that combining reference samples with foundation-model rewards enables distribution transfer and flexible style customization. In human evaluation, our method outperforms Flow-GRPO and SD3, achieving 70.0% and 72.4% win rates in image quality and aesthetics, respectively. Code and models have been released.
Read the full abstract
Computer Vision
XiCAD: Camera Activation Detection in the Da Vinci Xi User Interface
Purpose: Robot-assisted minimally invasive surgery relies on endoscopic video as the sole intraoperative visual feedback. The DaVinci Xi system overlays a graphical user interface (UI) that indicates the state of each robotic arm, including the activation of the endoscope arm. Detecting this activation provides valuable metadata such as camera movement information, which can support downstream surgical data science tasks including tool tracking, skill assessment, or camera control automation. Methods: We developed a lightweight pipeline based on a ResNet18 convolutional neural network to automatically identify the position of the camera tile and its activation state within the DaVinci Xi UI. The model was fine-tuned on manually annotated data from the SurgToolLoc dataset and evaluated across three public datasets comprising over 70,000 frames. Results: The model achieved F1-scores between 0.993 and 1.000 for the binary detection of active cameras and correctly localized the camera tile in all cases without false multiple-camera detections. Conclusion: The proposed pipeline enables reliable, real-time extraction of camera activation metadata from surgical videos, facilitating automated preprocessing and analysis for diverse downstream applications. All code, trained models, and annotations are publicly available.
Read the full abstract
Computer Vision
Zoo3D: Zero-Shot 3D Object Detection at Scene Level
3D object detection is fundamental for spatial understanding. Real-world environments demand models capable of recognizing diverse, previously unseen objects, which remains a major limitation of closed-set methods. Existing open-vocabulary 3D detectors relax annotation requirements but still depend on training scenes, either as point clouds or images. We take this a step further by introducing Zoo3D, the first training-free 3D object detection framework. Our method constructs 3D bounding boxes via graph clustering of 2D instance masks, then assigns semantic labels using a novel open-vocabulary module with best-view selection and view-consensus mask generation. Zoo3D operates in two modes: the zero-shot Zoo3D$_0$, which requires no training at all, and the self-supervised Zoo3D$_1$, which refines 3D box prediction by training a class-agnostic detector on Zoo3D$_0$-generated pseudo labels. Furthermore, we extend Zoo3D beyond point clouds to work directly with posed and even unposed images. Across ScanNet200 and ARKitScenes benchmarks, both Zoo3D$_0$ and Zoo3D$_1$ achieve state-of-the-art results in open-vocabulary 3D object detection. Remarkably, our zero-shot Zoo3D$_0$ outperforms all existing self-supervised methods, hence demonstrating the power and adaptability of training-free, off-the-shelf approaches for real-world 3D understanding. Code is available at https://github.com/col14m/zoo3d .
Read the full abstract
Computer Vision
Diversifying Long Prompt Image Generation through Structured Prompt Embedding Space Sampling
Modern text-to-image models produce impressive visual results from richly specified prompts, yet their behavior under long prompts remains insufficiently understood. In this paper, we study a practical failure mode in which accumulated semantic constraints progressively suppress output variation, causing diversity to collapse even when many visual factors remain unspecified. We show that this phenomenon appears consistently across recent generation models as prompt length increases, and provide a theoretical motivation that connects long-prompt conditioning with reduced sampling entropy in the prompt embedding space. Based on this observation, we introduce PromptMoG, a training-free approach that samples prompt embeddings from a Mixture-of-Gaussians distribution to restore generative flexibility while maintaining semantic fidelity. To support systematic evaluation, we further present LPD-Bench, a structured benchmark of long and semantically dense prompts for measuring both fidelity and diversity under compositional text conditioning. Extensive experiments on four large-scale diffusion models, including SD3.5-Large, Flux.1-Krea-Dev, CogView4, and Qwen-Image, show that PromptMoG consistently improves diversity for long-prompt image generation. The code is publicly available at https://github.com/basiclab/PromptMoG.
Read the full abstract
Computer Vision
Uplifting Table Tennis: A Robust, Real-World Application for 3D Trajectory and Spin Estimation
Obtaining the precise 3D motion of a table tennis ball from standard monocular videos is a challenging problem, as existing methods trained on synthetic data struggle to generalize to the noisy, imperfect ball and table detections of the real world. This is primarily due to the inherent lack of 3D ground truth trajectories and spin annotations for real-world video. To overcome this, we propose a novel two-stage pipeline that divides the problem into a front-end perception task and a back-end 2D-to-3D uplifting task. This separation allows us to train the front-end components with abundant 2D supervision from our newly created TTHQ dataset, while the back-end uplifting network is trained exclusively on physically-correct synthetic data. We specifically re-engineer the uplifting model to be robust to common real-world artifacts, such as missing detections and varying frame rates. By integrating a ball detector and a table keypoint detector, our approach transforms a proof-of-concept uplifting method into a practical, robust, and high-performing end-to-end application for 3D table tennis trajectory and spin analysis.
Read the full abstract
Computer Vision
HistoSpeckle-Net: Mutual Information-Guided Deep Learning for high-fidelity reconstruction of complex OrganAMNIST images via perturbed Multimode Fibers
Existing deep learning methods in multimode fiber (MMF) imaging often focus on simpler datasets, limiting their applicability to complex, real-world imaging tasks. These models are typically data-intensive, a challenge that becomes more pronounced when dealing with diverse and complex images. In this work, we propose HistoSpeckle-Net, a deep learning architecture designed to reconstruct structurally rich medical images from MMF speckles. To build a clinically relevant dataset, we develop an optical setup that couples laser light through a spatial light modulator (SLM) into an MMF, capturing output speckle patterns corresponding to input OrganAMNIST images. Unlike previous MMF imaging approaches, which have not considered the underlying statistics of speckles and reconstructed images, we introduce a distribution-aware learning strategy. We employ a histogram-based mutual information loss to enhance model robustness and reduce reliance on large datasets. Our model includes a histogram computation unit that estimates smooth marginal and joint histograms for calculating mutual information loss. It also incorporates a unique Three-Scale Feature Refinement Module, which leads to multiscale Structural Similarity Index Measure (SSIM) loss computation. Together, these two loss functions enhance both the structural fidelity and statistical alignment of the reconstructed images. Our experiments on the complex OrganAMNIST dataset demonstrate that HistoSpeckle-Net achieves higher fidelity than baseline models such as U-Net and Pix2Pix. It gives superior performance even with limited training samples and across varying fiber bending conditions. By effectively reconstructing complex anatomical features with reduced data and under fiber perturbations, HistoSpeckle-Net brings MMF imaging closer to practical deployment in real-world clinical environments.
Read the full abstract
Computer Vision
V-Attack: Targeting Disentangled Value Features for Controllable Adversarial Attacks on LVLMs
Adversarial attacks have evolved from simply disrupting predictions on conventional task-specific models to the more complex goal of manipulating image semantics on Large Vision-Language Models (LVLMs). However, existing methods struggle with controllability and fail to precisely manipulate the semantics of specific concepts in the image. We attribute this limitation to semantic entanglement in the patch-token representations on which adversarial attacks typically operate: global context aggregated by self-attention in the vision encoder dominates individual patch features, making them unreliable handles for precise local semantic manipulation. Our systematic investigation reveals a key insight: value features (V) computed within the transformer attention block serve as much more precise handles for manipulation. We show that V suppresses global-context channels, allowing it to retain high-entropy, disentangled local semantic information. Building on this discovery, we propose V-Attack, a novel method designed for precise local semantic attacks. V-Attack targets the value features and introduces two core components: (1) a Self-Value Enhancement module to refine V's intrinsic semantic richness, and (2) a Text-Guided Value Manipulation module that leverages text prompts to locate source concept and optimize it toward a target concept. By bypassing the entangled patch features, V-Attack achieves highly effective semantic control. Extensive experiments across diverse LVLMs, including LLaVA, InternVL, DeepseekVL and GPT-4o, show that V-Attack improves the attack success rate by an average of 36% over state-of-the-art methods, exposing critical vulnerabilities in modern visual-language understanding. Our code and data are available https://github.com/Summu77/V-Attack.
Read the full abstract
Computer Vision
Patch-Level Glioblastoma Subregion Classification with a Contrastive Learning-Based Encoder
The significant molecular and pathological heterogeneity of glioblastoma, an aggressive brain tumor, complicates diagnosis and patient stratification. While traditional histopathological assessment remains the standard, deep learning offers a promising path toward objective and automated analysis of whole slide images. For the BraTS-Path 2025 Challenge, we developed a method that fine-tunes a pre-trained Vision Transformer (ViT) encoder with a dedicated classification head on the official training dataset. Our model's performance on the online validation set, evaluated via the Synapse platform, yielded a Matthews Correlation Coefficient (MCC) of 0.7064 and an F1-score of 0.7676. On the final test set, the model achieved an MCC of 0.6509 and an F1-score of 0.5330, which secured our team second place in the BraTS-Pathology 2025 Challenge. Our results establish a solid baseline for ViT-based histopathological analysis, and future efforts will focus on bridging the performance gap observed on the unseen validation data.
Read the full abstract
Computer Vision
Text-guided Controllable Diffusion for Realistic Camouflage Images Generation
Camouflage Images Generation (CIG) is an emerging research area that focuses on synthesizing images in which objects are harmoniously blended and exhibit high visual consistency with their surroundings. Existing methods perform CIG by either fusing objects into specific backgrounds or outpainting the surroundings via foreground object-guided diffusion. However, they often fail to obtain natural results because they overlook the logical relationship between camouflaged objects and background environments. To address this issue, we propose CT-CIG, a Controllable Text-guided Camouflage Images Generation method that produces realistic and logically plausible camouflage images. Leveraging Large Visual Language Models (VLM), we design a Camouflage-Revealing Dialogue Mechanism (CRDM) to annotate existing camouflage datasets with high-quality text prompts. Subsequently, the constructed image-prompt pairs are utilized to finetune Stable Diffusion, incorporating a lightweight controller to guide the location and shape of camouflaged objects for enhanced camouflage scene fitness. Moreover, we design a Frequency Interaction Refinement Module (FIRM) to capture high-frequency texture features, facilitating the learning of complex camouflage patterns. Extensive experiments, including CLIPScore evaluation and camouflage effectiveness assessment, demonstrate the semantic alignment of our generated text prompts and CT-CIG's ability to produce photorealistic camouflage images.
Read the full abstract
Computer Vision · Cross-listed from cs.AI
CostNav: A Navigation Benchmark for Real-World Economic-Cost Evaluation of Physical AI Agents
Current navigation benchmarks focus on task success but do not capture the economic constraints essential for commercializing autonomous delivery systems. We introduce CostNav, an Economic Navigation Benchmark that evaluates physical AI agents on a cost-revenue and break-even analysis, pairing Isaac Sim's collision and cargo dynamics with industry-standard data such as Securities and Exchange Commission (SEC) filings and Abbreviated Injury Scale (AIS) injury reports. To our knowledge, CostNav is the first physics-grounded economic benchmark to use regulatory and financial data to quantify the gap between navigation metrics and commercial deployment, revealing that high task-success rates alone do not ensure economic viability. Evaluating seven baselines (two rule-based and five imitation-learning methods), we find no method economically viable: all yield negative contribution margins. CANVAS, using only an RGB camera and GPS, attains the highest task success and the least-negative margin among methods with non-zero Service-Level Agreement (SLA) compliance (-\$28.40/run), outperforming LiDAR-equipped Nav2 w/ GPS (-\$37.34/run). A sim-trained policy evaluated on a real delivery robot yields SLA compliance close to its simulation result, indicating that policy performance in CostNav's simulation transfers to real-world deployment. We challenge the community to achieve economic viability on CostNav, which scores methods by cost-revenue outcomes. All resources are available at https://github.com/worv-ai/CostNav.
Read the full abstract
Computer Vision
OmniAlpha: Aligning Transparency-Aware Generation via Multi-Task Unified Reinforcement Learning
Transparency-aware generation requires modeling not only RGB appearance but also alpha-based opacity and cross-layer composition, which are essential for tasks such as image matting, object removal, layer decomposition, and multi-layer content creation. However, existing RGBA-related methods remain largely fragmented, with separate pipelines designed for individual tasks. While a unified model is desirable, supervised fine-tuning alone is insufficient, as localized regression objectives cannot directly optimize the compositional fidelity, alpha-boundary precision, and structural consistency required for high-quality RGBA generation. To address this, we propose OmniAlpha, a unified multi-task reinforcement learning framework for transparency-aware generation and manipulation. OmniAlpha combines an end-to-end alpha-aware VAE and a sequence-to-sequence Diffusion Transformer, with a bi-directional layer axis in positional encoding to jointly model multiple RGBA inputs and outputs within a single forward pass. Built on a multi-task SFT cold start, it further performs GRPO-style post-training with layer-aware rewards defined on decoded RGBA outputs, enabling direct optimization of cross-layer coherence and fine transparency details. Experiments across five categories of transparency-aware tasks show that OmniAlpha consistently outperforms its unified SFT baseline and achieves strong performance against specialized expert models, including a 9.07% relative reduction in RGB L1 on layer decomposition and 74%/68% improvements over conventional matting tools on SAD/Grad for automatic matting.
Read the full abstract
Computer Vision
Robust 3D Brain MRI Inpainting with Random Masking Augmentation
The ASNR-MICCAI BraTS-Inpainting Challenge was established to mitigate dataset biases that limit deep learning models in the quantitative analysis of brain tumor MRI. This paper details our submission to the 2025 challenge, a novel deep learning framework for synthesizing healthy tissue in 3D scans. The core of our method is a U-Net architecture trained to inpaint synthetically corrupted regions, enhanced with a random masking augmentation strategy to improve generalization. Quantitative evaluation confirmed the efficacy of our approach, yielding an SSIM of 0.873$\pm$0.004, a PSNR of 24.996$\pm$4.694, and an MSE of 0.005$\pm$0.087 on the validation set. On the final online test set, our method achieved an SSIM of 0.919$\pm$0.088, a PSNR of 26.932$\pm$5.057, and an RMSE of 0.052$\pm$0.026. This performance secured first place in the BraTS-Inpainting 2025 challenge and surpassed the winning solutions from the 2023 and 2024 competitions on the official leaderboard.
Read the full abstract
Computer Vision
GHR-VQA: Graph-guided Hierarchical Relational Reasoning for Video Question Answering
We propose GHR-VQA, Graph-guided Hierarchical Relational Reasoning for Video Question Answering (Video QA), a novel human-centric framework that incorporates scene graphs to capture intricate human-object interactions within video sequences. Unlike traditional pixel-based methods, each frame is represented as a scene graph and human nodes across frames are linked to a global root, forming the video-level graph and enabling cross-frame reasoning centered on human actors. The video-level graphs are then processed by Graph Neural Networks (GNNs), transforming them into rich, context-aware embeddings for efficient processing. Finally, these embeddings are integrated with question features in a hierarchical network operating across different abstraction levels, enhancing both local and global understanding of video content. This explicit human-rooted structure enhances interpretability by decomposing actions into human-object interactions and enables a more profound understanding of spatiotemporal dynamics. We validate our approach on the Action Genome Question Answering (AGQA) dataset, achieving significant performance improvements, including a 7.3% improvement in object-relation reasoning over the state of the art.
Read the full abstract
Computer Vision
SFA: Scan, Focus, and Amplify toward Guidance-aware Answering for Video TextVQA
Video text-based visual question answering (Video TextVQA) task aims to answer questions about videos by leveraging the visual text appearing within the videos. This task poses significant challenges, requiring models to accurately perceive and comprehend scene text that varies in scale, orientation, and clarity across frames, while effectively integrating temporal and semantic context to generate precise answers. Moreover, the model must identify question-relevant textual cues and filter out redundant or irrelevant information to ensure answering is guided by the most relevant and informative cues. To address these challenges, we propose SFA, a training-free framework and the first Video-LLM-based method tailored for Video TextVQA, motivated by the human process of answering questions. By adaptively scanning video frames, selectively focusing on key regions, and directly amplifying them, SFA effectively guides the Video-LLM's attention toward essential cues, enabling it to generate more accurate answers. SFA achieves new state-of-the-art results across several public Video TextVQA datasets and surpasses previous methods by a substantial margin, demonstrating its effectiveness and generalizability.
Read the full abstract
Computer Vision
Exo2EgoSyn: Unlocking Foundation Video Generation Models for Exocentric-to-Egocentric Video Synthesis
Foundation video generation models such as WAN 2.2 exhibit strong text- and image-conditioned synthesis abilities but remain constrained to the same-view generation setting. In this work, we introduce Exo2EgoSyn, an adaptation of WAN 2.2 that unlocks Exocentric-to-Egocentric(Exo2Ego) cross-view video synthesis. Our framework consists of three key modules. Ego-Exo View Alignment(EgoExo-Align) enforces latent-space alignment between exocentric and egocentric first-frame representations, reorienting the generative space from the given exo view toward the ego view. Multi-view Exocentric Video Conditioning (MultiExoCon) aggregates multi-view exocentric videos into a unified conditioning signal, extending WAN2.2 beyond its vanilla single-image or text conditioning. Furthermore, Pose-Aware Latent Injection (PoseInj) injects relative exo-to-ego camera pose information into the latent state, guiding geometry-aware synthesis across viewpoints. Together, these modules enable high-fidelity ego view video generation from third-person observations without retraining from scratch. Experiments on ExoEgo4D validate that Exo2EgoSyn significantly improves Ego2Exo synthesis, paving the way for scalable cross-view video generation with foundation models. Source code and models will be released publicly.
Read the full abstract
Computer Vision
Realizing Fully-Integrated, Low-Power, Event-Based Pupil Tracking with Neuromorphic Hardware
Eye tracking is fundamental to numerous applications, yet achieving robust, high-frequency tracking with ultra-low power consumption remains challenging for wearable platforms. While event-based vision sensors offer microsecond resolution and sparse data streams, they have lacked fully integrated, low-power processing solutions capable of real-time inference. In this work, we present the first battery-powered, wearable pupil-center-tracking system with complete on-device integration, combining event-based sensing and neuromorphic processing on the commercially available Speck2f system-on-chip with lightweight coordinate decoding on a low-power microcontroller. Our solution features a novel uncertainty-quantifying spiking neural network with gated temporal decoding, optimized for strict memory and bandwidth constraints, complemented by systematic deployment mechanisms that bridge the reality gap. We validate our system on a new multi-user dataset and demonstrate a wearable prototype with dual neuromorphic devices achieving robust binocular pupil tracking at 100 Hz with an average power consumption below 5 mW per eye. Our work demonstrates that end-to-end neuromorphic computing enables practical, always-on eye tracking for next-generation energy-efficient wearable systems.
Read the full abstract
Computer Vision
ADNet: A Large-Scale and Extensible Multi-Domain Benchmark for Anomaly Detection Across 380 Real-World Categories
Anomaly detection (AD) aims to identify defects using normal-only training data. Existing anomaly detection benchmarks (e.g., MVTec-AD with 15 categories) cover only a narrow range of categories, limiting the evaluation of cross-context generalization and scalability. We introduce ADNet, a large-scale, multi-domain benchmark comprising 380 categories aggregated from 49 publicly available datasets across Electronics, Industry, Agrifood, Infrastructure, and Medical domains. The benchmark includes a total of 196,294 RGB images, consisting of 116,192 normal samples for training and 80,102 test images, of which 60,311 are anomalous. All images are standardized with MVTec-style pixel-level annotations and structured text descriptions spanning both spatial and visual attributes, enabling multimodal anomaly detection tasks. Extensive experiments reveal a clear scalability challenge: existing state-of-the-art methods achieve 90.6% I-AUROC in one-for-one settings but drop to 78.5% when scaling to all 380 categories in a multi-class setting. To address this, we propose Dinomaly-m, a context-guided Mixture-of-Experts extension of Dinomaly that expands decoder capacity without increasing inference cost. It achieves 83.2% I-AUROC and 93.1% P-AUROC, demonstrating superior performance over existing approaches. ADNet is designed as a standardized and extensible benchmark, supporting the community in expanding anomaly detection datasets across diverse domains and providing a scalable foundation for future anomaly detection foundation models. Dataset: https://grainnet.github.io/ADNet
Read the full abstract
Computer Vision
Action Without Interaction: Probing the Physical Foundations of Video LMMs via Contact-Release Detection
Large multi-modal models (LMMs) show increasing performance in realistic visual tasks for images and, more recently, for videos. For example, given a video sequence, such models are able to describe in detail objects, the surroundings and dynamic actions. In this study, we explored the extent to which these models ground their semantic understanding in the actual visual input. Specifically, given sequences of hands interacting with objects, we asked models when and where the interaction begins or ends. For this purpose, we introduce a first of its kind, large-scale dataset with more than 20K annotated interactions on videos from the Something-Something-V2 dataset. 250 AMTurk human annotators labeled core interaction events, particularly when and where objects and agents become attached (`contact') or detached (`release'). We asked SoTA LMMs, including GPT, Gemini and Qwen to locate these events in short videos, each with a single event. The results show that while models reliably name target objects and identify actions, they exhibit a form of `shortcut learning' where semantic success masks a failure in physical grounding. Specifically, they consistently fail to identify the frame where the interaction begins or ends and poorly localize the physical event within the scene. This disconnect suggests that while LMMs excel at System 1 intuitive pattern recognition (naming the action and objects), they lack the System 2 cognitive foundations required to reason about physical primitives like `contact' and `release', hence truly ground dynamic scenes in physical reality.
Read the full abstract
Computer Vision
Harmonious Parameter Adaptation in Continual Visual Instruction Tuning for Safety-Aligned MLLMs
While continual visual instruction tuning (CVIT) has shown promise in adapting multimodal large language models (MLLMs), existing studies predominantly focus on models without safety alignment. This critical oversight ignores the fact that real-world MLLMs inherently require such mechanisms to mitigate potential risks. In this work, we shift our focus to CVIT for safety-aligned MLLMs and observe that during continual adaptation, the model not only suffers from task forgetting but also exhibits degradation in its safety. Achieving a harmonious balance between safety and task performance remains a crucial challenge. To address this, we propose Harmonious Parameter Adaptation (HPA), a post-training framework composed of focusing-based parameter partition, harmoniously balanced parameter selection, and orthogonal parameter adjustment. Specifically, HPA partitions parameters into two types based on their focus on safety or task performance, and selects the focused ones to preserve from a balanced perspective. In addition, HPA imposes orthogonality constraints on parameter updates to further alleviate catastrophic forgetting. Extensive experiments on the CVIT benchmark and safety evaluation datasets demonstrate that HPA better maintains high safety and mitigates forgetting than existing baselines. Code is available at https://github.com/Minato-Zackie/HPA.
Read the full abstract
Computer Vision
SKEL-CF: Coarse-to-Fine Biomechanical Skeleton and Surface Mesh Recovery
Parametric 3D human models such as SMPL have driven significant advances in human pose and shape estimation, yet their simplified kinematics limit biomechanical realism. The recently proposed SKEL model addresses this limitation by re-rigging SMPL with an anatomically accurate skeleton. However, estimating SKEL parameters directly remains challenging due to limited training data, perspective ambiguities, and the inherent complexity of human articulation. We introduce SKEL-CF, a coarse-to-fine framework for SKEL parameter estimation. SKEL-CF employs a transformer-based encoder-decoder architecture, where the encoder predicts coarse camera and SKEL parameters, and the decoder progressively refines them in successive layers. To ensure anatomically consistent supervision, we convert the existing SMPL-based dataset 4DHuman into a SKEL-aligned version, 4DHuman-SKEL, providing high-quality training data for SKEL estimation. In addition, to mitigate depth and scale ambiguities, we explicitly incorporate camera modeling into the SKEL-CF pipeline and demonstrate its importance across diverse viewpoints. Extensive experiments validate the effectiveness of the proposed design. On the challenging MOYO dataset, SKEL-CF achieves 85.0 MPJPE / 51.4 PA-MPJPE, significantly outperforming the previous SKEL-based state-of-the-art HSMR (104.5 / 79.6). These results establish SKEL-CF as a scalable and anatomically faithful framework for human motion analysis, facilitating the use of computer vision techniques in biomechanics-related analysis. Our implementation is available on the project page: https://pokerman8.github.io/SKEL-CF/.
Read the full abstract
Computer Vision
Map-World: Masked Action planning and Path-Integral World Model for Autonomous Driving
Motion planning for autonomous driving must handle multiple plausible futures while remaining computationally efficient. Recent end-to-end systems and world-model-based planners predict rich multi-modal trajectories, but typically rely on handcrafted anchors or reinforcement learning to select a single best mode for training and control. This selection discards information about alternative futures and complicates optimization. We propose MAP-World, a prior-free multi-modal planning framework that couples masked action planning with a path-weighted world model. The Masked Action Planning (MAP) module treats future ego motion as masked sequence completion: past waypoints are encoded as visible tokens, future waypoints are represented as mask tokens, and a driving-intent path provides a coarse scaffold. A compact latent planning state is expanded into multiple trajectory queries with injected noise, yielding diverse, temporally consistent modes without anchor libraries or teacher policies. A lightweight world model then rolls out future BEV semantics conditioned on each candidate trajectory. During training, semantic losses are computed as an expectation over modes, using trajectory probabilities as discrete path weights, so the planner learns from the full distribution of plausible futures instead of a single selected path. On NAVSIM, our method matches anchor-based approaches and achieves state-of-the-art performance among world-model-based methods, while avoiding reinforcement learning and maintaining real-time inference latency.
Read the full abstract
Computer Vision
Alzheimers Disease Progression Prediction Based on Manifold Mapping of Irregularly Sampled Longitudinal Data
The uncertainty of clinical examinations frequently leads to irregular observation intervals in longitudinal imaging data, posing challenges for modeling disease progression.Most existing imaging-based disease prediction models operate in Euclidean space, which assumes a flat representation of data and fails to fully capture the intrinsic continuity and nonlinear geometric structure of irregularly sampled longitudinal images. To address the challenge of modeling Alzheimers disease (AD) progression from irregularly sampled longitudinal structural Magnetic Resonance Imaging (sMRI) data, we propose a Riemannian manifold mapping, a Time-aware manifold Neural ordinary differential equation, and an Attention-based riemannian Gated recurrent unit (R-TNAG) framework. Our approach first projects features extracted from high-dimensional sMRI into a manifold space to preserve the intrinsic geometry of disease progression. On this representation, a time-aware Neural Ordinary Differential Equation (TNODE) models the continuous evolution of latent states between observations, while an Attention-based Riemannian Gated Recurrent Unit (ARGRU) adaptively integrates historical and current information to handle irregular intervals. This joint design improves temporal consistency and yields robust AD trajectory prediction under irregular sampling.Experimental results demonstrate that the proposed method consistently outperforms state-of-the-art models in both disease status prediction and cognitive score regression. Ablation studies verify the contributions of each module, highlighting their complementary roles in enhancing predictive accuracy. Moreover, the model exhibits stable performance across varying sequence lengths and missing data rates, indicating strong temporal generalizability. Cross-dataset validation further confirms its robustness and applicability in diverse clinical settings.
Read the full abstract
Computer Vision
Restora-Flow: Mask-Guided Image Restoration with Flow Matching
Flow matching has emerged as a promising generative approach that addresses the lengthy sampling times associated with state-of-the-art diffusion models and enables a more flexible trajectory design, while maintaining high-quality image generation. This capability makes it suitable as a generative prior for image restoration tasks. Although current methods leveraging flow models have shown promising results in restoration, some still suffer from long processing times or produce over-smoothed results. To address these challenges, we introduce Restora-Flow, a training-free method that guides flow matching sampling by a degradation mask and incorporates a trajectory correction mechanism to enforce consistency with degraded inputs. We evaluate our approach on both natural and medical datasets across several image restoration tasks involving a mask-based degradation, i.e., inpainting, super-resolution and denoising. We show superior perceptual quality and processing time compared to diffusion and flow matching-based reference methods.
Read the full abstract
Computer Vision
A Compact Hybrid Convolution–Frequency State Space Network for Learned Image Compression
Learned image compression (LIC) has recently benefited from Transformer- and state space models (SSM)- based backbones for modeling long-range dependencies. However, the former typically incurs quadratic complexity, whereas the latter often disrupts neighborhood continuity by flattening 2D features into 1D sequences. To address these issues, we propose a compact Hybrid Convolution and Frequency State Space Network (HCFSSNet) for LIC. HCFSSNet combines convolutional layers for local detail modeling with a Vision Frequency State Space (VFSS) block for complementary long-range contextual aggregation. Specifically, the VFSS block consists of a Vision Omni-directional Neighborhood State Space (VONSS) module, which scans features along horizontal, vertical, and diagonal directions to better preserve 2D neighborhood relations, and an Adaptive Frequency Modulation Module (AFMM), which performs discrete cosine transform-based adaptive reweighting of frequency components. In addition, we introduce a Frequency Swin Transformer Attention Module (FSTAM) in the hyperprior path to enhance frequency-aware side information modeling. Experiments on the benchmark datasets show that the proposed HCFSSNet achieves a competitive rate-distortion performance against recent LIC codecs. The source code and models will be made publicly available.
Read the full abstract
Computer Vision
Vision-Language Models for Automated 3D PET/CT Report Generation
Positron emission tomography/computed tomography (PET/CT) is essential in oncology, yet the rapid expansion of scanners has outpaced the availability of trained specialists, making automated PET/CT report generation (PETRG) increasingly important for reducing clinical workload. Compared with structural imaging (e.g., X-ray, CT, and MRI), functional PET poses distinct challenges: metabolic patterns vary with tracer physiology, and whole-body 3D contextual information is required rather than local-region interpretation. To advance PETRG, we propose PETRG-3D, an end-to-end 3D dual-branch framework that separately encodes PET and CT volumes and incorporates style-adaptive prompts to mitigate inter-hospital variability in reporting practices. We construct PETRG-Lym, a multi-center lymphoma dataset collected from four hospitals (824 reports w/ 245,509 paired PET/CT slices), and construct AutoPET-RG-Lym, a publicly accessible PETRG benchmark derived from open imaging data but equipped with new expert-written, clinically validated reports (135 cases). To assess clinical utility, we introduce PETRG-Score, a lymphoma-specific evaluation protocol that jointly measures metabolic and structural findings across curated anatomical regions. Experiments show that PETRG-3D substantially outperforms existing methods on both natural language metrics (e.g., +31.49% ROUGE-L) and clinical efficacy metrics (e.g., +8.18% PET-All), highlighting the benefits of volumetric dual-modality modeling and style-aware prompting. Overall, this work establishes a foundation for future PET/CT-specific models emphasizing disease-aware reasoning and clinically reliable evaluation. Codes, models, and AutoPET-RG-Lym will be released.
Read the full abstract
Computer Vision
UltraViCo: Breaking Extrapolation Limits in Video Diffusion Transformers
Despite advances, video diffusion transformers still struggle to generalize beyond their training length, a challenge we term video length extrapolation. We identify two failure modes: model-specific periodic content repetition and a universal quality degradation. Prior works attempt to solve repetition via positional encodings, overlooking quality degradation and achieving only limited extrapolation. In this paper, we revisit this challenge from a more fundamental view: attention maps, which directly govern how context influences outputs. We identify that both failure modes arise from a unified cause: attention dispersion, where tokens beyond the training window dilute learned attention patterns. This leads to quality degradation and repetition emerges as a special case when this dispersion becomes structured into periodic attention patterns, induced by harmonic properties of positional encodings. Building on this insight, we propose UltraViCo, a training-free, plug-and-play method that suppresses attention for tokens beyond the training window via a constant decay factor. By jointly addressing both failure modes, we outperform a broad set of baselines largely across models and extrapolation ratios, pushing the extrapolation limit from 2x to 4x. Remarkably, it improves Dynamic Degree and Imaging Quality by 233% and 40.5% over the previous best method at 4x extrapolation. Furthermore, our method generalizes seamlessly to downstream tasks such as controllable video synthesis and editing.
Read the full abstract
Computer Vision
LungEvaty: A Scalable, Open-Source Transformer-based Deep Learning Model for Lung Cancer Risk Prediction in LDCT Screening
Lung cancer risk estimation is gaining increasing importance as more countries introduce population-wide screening programs using low-dose CT (LDCT). As imaging volumes grow, scalable methods that can process entire lung volumes efficiently are essential to tap into the full potential of these large screening datasets. Existing approaches either over-rely on pixel-level annotations, limiting scalability, or analyze the lung in fragments, weakening performance. We present LungEvaty, a fully transformer-based framework for predicting 1-6 year lung cancer risk from a single LDCT scan. The model operates on whole-lung inputs, learning directly from large-scale screening data to capture comprehensive anatomical and pathological cues relevant for malignancy risk. Using only imaging data and no region supervision, LungEvaty matches state-of-the-art performance, refinable by an optional Anatomically Informed Attention Guidance (AIAG) loss that encourages anatomically focused attention. In total, LungEvaty was trained on more than 90,000 CT scans, including over 28,000 for fine-tuning and 6,000 for evaluation. The framework offers a simple, data-efficient, and fully open-source solution that provides an extensible foundation for future research in longitudinal and multimodal lung cancer risk prediction.
Read the full abstract
Computer Vision
Multi Head Attention Enhanced Inception v3 for Cardiomegaly Detection
The healthcare industry has been revolutionized significantly by novel imaging technologies, not just in the diagnosis of cardiovascular diseases but also by the visualization of structural abnormalities like cardiomegaly. This article explains an integrated approach to the use of deep learning tools and attention mechanisms for automatic detection of cardiomegaly using X-ray images. The initiation of the project is grounded on a strong Data Collection phase and gathering the data of annotated X-ray images of various types. Then, while the Preprocessing module fine-tunes image quality, it is feasible to utilize the best out of the data quality in the proposed system. In our proposed system, the process is a CNN configuration leveraging the inception V3 model as one of the key blocks. Besides, we also employ a multilayer attention mechanism to enhance the strength. The most important feature of the method is the multi-head attention mechanism that can learn features automatically. By exact selective focusing on only some regions of input, the model can thus identify cardiomegaly in a sensitive manner. Attention rating is calculated, duplicated, and applied to enhance representation of main data, and therefore there is a successful diagnosis. The Evaluation stage will be extremely strict and it will thoroughly evaluate the model based on such measures as accuracy and precision. This will validate that the model can identify cardiomegaly and will also show the clinical significance of this method. The model has accuracy of 95.6, precision of 95.2, recall of 96.2, sensitivity of 95.7, specificity of 96.1 and an Area Under Curve(AUC) of 96.0 and their respective graphs are plotted for visualisation.
Read the full abstract
Computer Vision
Exploring State-of-the-art models for Early Detection of Forest Fires
There have been many recent developments in the use of Deep Learning Neural Networks for fire detection. In this paper, we explore an early warning system for detection of forest fires. Due to the lack of sizeable datasets and models tuned for this task, existing methods suffer from missed detection. In this work, we first propose a dataset for early identification of forest fires through visual analysis. Unlike existing image corpuses that contain images of wide-spread fire, our dataset consists of multiple instances of smoke plumes and fire that indicates the initiation of fire. We obtained this dataset synthetically by utilising game simulators such as Red Dead Redemption 2. We also combined our dataset with already published images to obtain a more comprehensive set. Finally, we compared image classification and localisation methods on the proposed dataset. More specifically we used YOLOv7 (You Only Look Once) and different models of detection transformer.