In this work, we present WeatherEdit, a novel weather editing pipeline for generating realistic weather effects with controllable types and severity in 3D scenes. Our approach is structured into two key components: weather background editing and weather particle construction. For weather background editing, we introduce an all-in-one adapter that integrates multiple weather styles into a single diffusion model, enabling the generation of diverse weather effects in 2D image backgrounds. During inference, we design a Temporal-View (TV-) attention mechanism that follows a specific order to aggregate temporal and spatial information, ensuring consistent editing across multi-frame and multi-view images. To construct the weather particles, we first reconstruct a 3D scene using the edited images and then introduce a 4D Gaussian field to generate snowflakes, raindrops and fog in the scene. The attributes and dynamics of these particles are controlled through attribute modelling and dynamic simulation, ensuring realistic weather representation and flexible severity adjustments. Finally, we integrate the 4D Gaussian field with the 3D scene to render consistent and highly realistic weather effects. Experiments on multiple driving datasets demonstrate that WeatherEdit can generate diverse weather effects with controllable condition severity, highlighting its potential for autonomous driving simulation in adverse weather.
论文检索
输入标题、作者或关键词，从 100,903 篇学术成果中精准定位
Scene text image super-resolution aims to enhance text legibility by recovering high-resolution text images from low-resolution inputs. However, maintaining fine details such as text strokes, edges, and textual accuracy remains challenging, particularly in low-light environments and high-speed motion scenarios, where degradation is more severe. Event cameras, with their high temporal resolution and ability to capture intensity changes, offer a promising solution for restoring lost fine details and mitigating degradation in these challenging conditions. In this paper, we propose EvTSR, the first framework that integrates Event data for scene Text image Super-Resolution. The core of EvTSR is the dual-stream frequency boost (DSFB) mechanism, which separates image features into high- and low-frequency components. High-frequency details like edges and strokes are enhanced using event data via the event-guided high-frequency (EGH) mechanism, while low-frequency components, responsible for global structure, are refined using the Text-Guided Low-frequency (TGL) mechanism with a pre-trained text recognizer, ensuring textual coherence. To further improve cross-modal integration, we introduce the cross-modal fusion (CMF) mechanism, which effectively aligns event and image features, enabling robust information fusion. Extensive experiments demonstrate that EvTSR achieves superior performance over existing methods.
Game Ground Bench: Probing the Limits of LVLMs in Complex Semantic Grounding Across Game Universes
PDF ↗Large Vision-Language Models (LVLMs) have demonstrated remarkable capabilities, yet their ability to ground language in complex, interactive environments such as video games remains a critical frontier. Existing benchmarks are inadequate for this purpose: real-world datasets like RefCOCO introduce a domain gap; GUI-centric benchmarks lack the complexity of modern game interfaces; and existing game-specific benchmarks are often too simplistic or narrow, failing to assess fine-grained, generalizable grounding capabilities. To address this issue, we propose GGBench — a large-scale, cross-genre benchmark designed to probe the grounding capabilities of LVLMs in diverse gaming scenarios. GGBench features unprecedented genre diversity, encompassing 10 categories including card games, first-person shooters, and role-playing games, with a total of 1335 test images. It focuses on tasks that require connecting natural language instructions to specific in-game objects and UI elements. Experimental results show existing models perform poorly on GGBench, with weak grounding abilities, especially in complex game scenarios. Due to limited data scale, fine-tuning them for gaming scenarios is also challenging. To address this, we propose Game-R1, a novel training method centered on the Grounded Reinforcement Policy Optimization (GRPO) algorithm. GRPO maximizes limited interaction data utility and enables robust few-shot generalization across games. Extensive experiments show Game-R1 significantly outperforms existing LVLMs on GGBench, validating our approach. GGBench provides a solid and comprehensive evaluation platform for subsequent research on agents in gaming environments, which strongly promotes development in this field.
What to Trust? A Trust-aware Knowledge-guided Method for Zero-shot Object State Understanding in Videos
PDF ↗Object state understanding aims at recognizing the co-occurrence and transitions of multiple object states in videos. While learning from videos handles seen object states well, it struggles with novel ones. We address this task in a zero-shot setting by extracting state-specific knowledge from pre-trained models and using Vision-Language Models (VLMs) to verify whether such knowledge is visually grounded in videos. However, the extracted knowledge varies in its ability to distinguish states, and VLM observations are not always trustworthy. To address this issue, we propose a trust-aware knowledge-guided method to model knowledge trustworthiness and emphasize highly discriminative knowledge that VLMs can reliably observe. Specifically, we collect spatial knowledge for each object state from retrieved images and cues generated from a Large Language Model, then use VLMs to vote on each knowledge element by scoring its visual consistency with the video. In addition to a single scene, temporal dependencies of object states across scenes are also captured using a generative VLM. Under spatial and temporal constraints, we propose an adaptive knowledge refinement module that iteratively updates knowledge reliability weights to achieve a global consensus in object state inference across the video. Finally, object states are inferred by combining the refined weights with VLM voting results. Experiments on two datasets demonstrate the effectiveness of our method.
Video object detection is a fundamental yet challenging task in computer vision. Recently, DETR-based methods have gained prominence in this domain owing to their powerful global modeling capabilities. However, these methods are still confronted with two key limitations: frame-agnostic initialization of object queries and scale-agnostic attention mechanisms, which hinder their capability to capture the appearance variations of dynamic objects and model the temporal consistency across frames. To alleviate these limitations, we propose a multiscale-aware transformer diffusion network (MSTDiff), a novel framework designed for the video object detection task, including two technical improvements over existing methods. First, we design a diffusion-driven adaptive query module, which models the object query distribution through a diffusion process conditioned on input frames, enabling an adaptive and content-aware initialization of object queries. Second, we develop a multiscale-aware transformer encoder module, which combines multi-head convolutional units with attention mechanisms to enhance multi-scale feature representations while preserving global dependence modeling. We conduct extensive experiments on the public ImageNet VID dataset, and the results demonstrate that our MSTDiff achieves 87.7% mAP with ResNet-101, outperforming most previous state-of-the-art video object detection methods.
Domain-adaptive person search (DAPS) aims to transfer pedestrian detection and re-identification capabilities from a labeled source domain to an unlabeled target domain, yet faces critical challenges from domain shift: semantic confusion among overlapping instances, over-reliance on shallow features for look-alike targets, and poor discriminability of small-scale instances. To address these issues, we propose the Localization-Anchored Instance Discrimination (LAID) framework, which leverages spatial relationships between bounding boxes as auxiliary signals to enhance instance identity learning. LAID integrates three complementary strategies: 1) Cost-Aware Instance Matching (CAIM) uses IoU-based global optimal assignment to align current detections with historical identities, reducing overlap-induced misassociations; 2) Dual-Scope Contrastive Learning (DSCL) combines spatial separation constraints (for geometrically distant pairs) with global contrastive learning, prompting the model to learn deep discriminative features beyond superficial similarities; 3) Task-Sensitivity Alignment (TSA) aligns confidence distributions of detection and ReID heads via KL divergence, ensuring consistent pseudo-label generation. Extensive experiments on CUHK-SYSU and PRW datasets demonstrate that LAID outperforms state-of-the-art DAPS methods, validating its effectiveness in mitigating domain shift and narrowing the performance gap between supervised and domain-adaptive person search.
Unsupervised Domain Adaptation (UDA) is a challenging task in person search. It adapts a well-trained model from a labeled source domain to an unlabeled target domain for privacy and efficiency. Currently, most of the state-of-the-art UDA person search methods adopt multi-scale feature alignment techniques to learn domain-invariant representations. However, person search is a multi-granularity task, and such an indiscriminate method of bridging the differences between domains misleads the identity learning process, which significantly limits the model's performance. In this paper, we propose an Instance-Guided Scene Adaptation (IGSA) framework by eradicating scene disparities and focusing the tasks on instances, effectively eliminating the contradiction between person search and domain adaptation. In IGSA, a Scene-Aware Bidirectional Filter (SABF) is designed to divide the image features into background and foreground to perform bidirectional modulations, thereby achieving simultaneous scene elimination and instance enhancement. To further improve the reliability of identity learning, we also propose an Instance Consistency Contrastive Learning (ICCL) method. By performing cross-epoch updates on the instance-level memory bank and re-initializing the cluster-level memory bank, the problem of inconsistent training across epochs caused by instance identity drift can be alleviated. Through the above designs, our method can achieve state-of-the-art performance on two benchmark datasets, with 82.1% mAP and 83.8% top-1 on the CUHK-SYSU dataset and 41.1% mAP and 82.3% top-1 on the PRW dataset, which is even better than some supervised methods.
Fetal ultrasound screening is a uniquely complex diagnostic task involving the simultaneous assessment of multiple fetal organs—each with its own anatomical and clinical context—within a single examination. Automating report generation for such cases poses a significant challenge: unlike existing methods that focus on single-organ radiology tasks (e.g., chest X-rays), fetal ultrasound requires reasoning over a structured, multiple-to-multiple setting, i.e., multi-organ images corresponding to a multi-section report. In this paper, we introduce FetusR, the first large-scale dataset for multi-organ fetal ultrasound reporting, containing 15,594 real-world cases with rich organ-wise annotations. To address the intrinsic image-report alignment, we propose Organ-Aware Routing Mixture-of-Retrieval Augmented Generation (ORM-RAG) inspired by the Mixture-of-Experts paradigm. Our method decomposes the complex alignment problem into multiple one-to-one sub-retrieval tasks. Specifically, ORM-RAG integrates (1) an organ-aware mixture-of-retrieval module that partitions the retrieval space into organ-specific corpora for independent retrieval, and (2) a dynamic routing mechanism that selectively aggregates high-confidence organ-specific reports while filtering uncertain ones. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art baselines across both textual similarity and clinical accuracy metrics. Our work opens a new direction for long-form, structured report generation in real-world, multi-organ medical imaging scenarios.
Unified Mixture-of-Experts Framework for Joint Cardiac and Vascular Ultrasound Analysis and Report Generation
PDF ↗Echocardiography and vascular ultrasound are essential for comprehensive cardiovascular assessment, yet manual evaluation and writing reports are labor-intensive, time-consuming, and require expertise from both cardiology and vascular surgery departments. Current automated report generation systems mainly focus on X-ray or CT, often neglecting echocardiographic modalities and critical quantitative parameters like aortic diameter and main pulmonary artery diameter, limiting their clinical utility. Moreover, the interdependence between cardiac and peripheral vascular health necessitates cross-departmental insights, which existing methods fail to incorporate. To address these limitations, we first propose the vision-language framework named the Echo-Cardiac-Vascular (ECV), for joint cardiac and vascular ultrasound report generation and parameter measurements. ECV introduces a Mixture-of-Experts vision encoder tailored for distinct ultrasound subtypes, a structured parameter measurement module for accurate quantification, and task-specific decoders that generate interpretable, multimodal diagnostic reports. Our framework, trained on 10K+ paired records, achieves high accuracy, improving diagnostic efficiency, consistency, and cross-disciplinary clinical applicability.
Recently, Test-Time Adaptation (TTA) has gained increasing attention in medical imaging due to its ability to improve model generalization under domain shifts without retraining. In particular, directly applying a well-trained model across various medical centers faces significant performance degradation caused by variations in equipment, operators, imaging conditions, and scanning skill levels of sonographers. Existing TTA methods either rely on parameter adaptation that increases computational cost or apply simple prediction fusion that ignores anatomical structure knowledge. To address these limitations, we propose a novel backward-free Topology-aware TTA framework named T^3 that integrates Structural Perception Modeling (SPM) and Box Regression Adaptation (BRA). SPM is implemented through an organ space heatmap generated via Gaussian kernel superposition. This heatmap encodes anatomical topology without requiring additional training or source data. BRA further improves localization and classification by fusing detection outputs based on the contribution of detected results to anatomically meaningful peak points from the heatmaps. Extensive experiments were conducted across six cross-domain scenarios, and the results demonstrate that our method achieves state-of-the-art cross-domain detection performance while maintaining high efficiency, offering a practical and robust solution for real-world medical diagnostic applications.
Fractured Glass, Failing Cameras: Simulating Physics-Based Adversarial Samples for Autonomous Driving Systems
PDF ↗While much research has recently focused on generating physics-based adversarial samples, a critical yet often overlooked category originates from physical failures within on-board cameras—components essential to the perception systems of autonomous vehicles. Camera failures, whether due to external stresses causing hardware breakdown or internal component faults, can directly jeopardize the safety and reliability of autonomous driving systems. Firstly, we motivate the study using two separate real-world experiments to showcase that indeed glass failures would cause the detection based neural network models to fail. Secondly, we develop a simulation-based study using the physical process of the glass breakage to create perturbed scenarios, representing a realistic class of physics-based adversarial samples. Using a finite element model (FEM)-based approach, we generate surface cracks on the camera image by applying a stress field defined by particles within a triangular mesh. Lastly, we use physically-based rendering (PBR) techniques to provide realistic visualizations of these physically plausible fractures. To assess the safety implications, we apply the simulated broken glass effects as image filters to two autonomous driving datasets- KITTI and BDD100K- as well as the large-scale image detection dataset MS-COCO. We then evaluate detection failure rates for critical object classes using CNN-based object detection models (YOLOv8 and Faster R-CNN) and a transformer-based architecture with Pyramid Vision Transformers. To further investigate the distributional impact of these visual distortions, we compute the Kullback-Leibler (K-L) divergence between three distinct data distributions, applying various broken glass filters to a custom dataset (captured through a cracked windshield), as well as the KITTI and Kaggle cats and dogs datasets. The K-L divergence analysis suggests that these broken glass filters do not introduce significant distributional shifts. Our goal is to provide a robust, physics-based methodology for generating adversarial samples that reflect real-world camera failures, with the overarching aim of improving the resilience and safety of autonomous driving systems against such physical threats.
Developing a multi-modal language model capable of understanding 3D scenes remains challenging due to the limited availability of 3D training data, in contrast to the abundance of 2D datasets used for vision-language models (VLMs). As an alternative, we introduce LLaVA³ (pronounced LLaVA Cube), a novel method that improves the 3D scene understanding capabilities of VLMs using only multi-view 2D images, and without requiring any fine-tuning. Inspired by Cubist painters, who represented multiple viewpoints of a 3D object within a single 2D picture, we propose to describe the 3D scene for the VLM through omnidirectional visual representations of each object. These representations are derived from an intermediate multi-view 3D reconstruction of the scene. Extensive experiments on 3D visual question answering and 3D language grounding show that our approach significantly outperforms previous 2D-based VLM solutions.
Video Anomaly Understanding (VAU) is a novel task focused on describing unusual occurrences in videos. Despite growing interest, the evaluation of VAU remains an open challenge. Existing benchmarks rely on n-gram-based metrics (e.g., BLEU, ROUGE-L) or LLM-based evaluation. The first fails to capture the rich, free-form, and visually grounded nature of LVLM responses, while the latter focuses on assessing language quality over factual relevance, often resulting in subjective judgments that are misaligned with human perception. In this work, we address this issue by proposing FineVAU, a new benchmark for VAU that shifts the focus towards rich, fine-grained and domain-specific understanding of anomalous videos. We formulate VAU as a three-fold problem, with the goal of comprehensively understanding key descriptive elements of anomalies in video: events (What), participating entities (Who) and location (Where). Our benchmark introduces a) FV-Score, a novel, human-aligned evaluation metric that assesses the presence of critical visual elements in LVLM answers, providing interpretable, fine-grained feedback; and b) FineW³, a novel, comprehensive dataset curated through a structured and fully automatic procedure that augments existing human annotations with high quality, fine-grained visual information. Human evaluation reveals that our proposed metric has a superior alignment with human perception of anomalies in comparison to current approaches. Detailed experiments on FineVAU unveil critical limitations in LVLM's ability to perceive anomalous events that require spatial and fine-grained temporal understanding, despite strong performance on coarse grain, static information, and events with strong visual cues.
Open-vocabulary semantic segmentation aims to assign pixel-level labels to images based on textual descriptions, even for categories beyond predefined closed sets. While vision-language foundation models like CLIP are widely used for this task, fine-tuning them for pixel-level predictions often compromises their generalization capabilities. To address this, we propose a novel fine-tuning strategy, CP-CLIP, which generates customized parameters for CLIP without sacrificing its generalization. Our method employs a customized parameter generator that produces newly added parameters based on random noise, using local visual features from CLIP's image encoder as conditions, enabling generalization to new images from unseen scenarios. Additionally, we introduce an orthogonal adaptation technique to ensure the update direction is orthogonal to the pre-trained weights, largely preserving the initial generalization ability. Extensive experiments demonstrate that CP-CLIP achieves state-of-the-art performance across multiple benchmarks in open-vocabulary semantic segmentation.
Personalized text-to-image generation aims to synthesize novel images of a specific subject or style using only a few reference images. Recent methods based on Low-Rank Adaptation (LoRA) enable efficient single-concept customization by injecting lightweight, concept-specific adapters into pre-trained diffusion models. However, combining multiple LoRA modules for multi-concept generation often leads to identity missing and visual feature leakage. In this work, we identify two key issues behind these failures: (1) token-wise interference among different LoRA modules, and (2) spatial misalignment between the attention map of a rare token and its corresponding concept-specific region. To address these issues, we propose Token-Aware LoRA (TARA), which introduces a token mask to explicitly constrain each module to focus on its associated rare token to avoid interference, and a training objective that encourages the spatial attention of a rare token to align with its concept region. Our method enables training-free multi-concept composition by directly injecting multiple independently trained TARA modules at inference time. Experimental results demonstrate that TARA enables efficient multi-concept inference and effectively preserving the visual identity of each concept by avoiding mutual interference between LoRA modules.
3D open-vocabulary scene understanding, which accurately perceives complex semantic properties of objects in space, has gained significant attention in recent years. In this paper, we propose GAGS, a framework that distills 2D CLIP features into 3D Gaussian splatting, enabling open-vocabulary queries for renderings on arbitrary viewpoints. The main challenge of distilling 2D features for 3D fields lies in the multiview inconsistency of extracted 2D features, which provides unstable supervision for the 3D feature field. GAGS addresses this challenge with two novel strategies. First, GAGS associates the prompt point density of SAM with the camera distances to scene objects, which significantly improves the multiview consistency of segmentation results. Second, GAGS further decodes a granularity factor to guide the distillation process and this granularity factor can be learned in a unsupervised manner to only select the multiview consistent 2D features in the distillation process. Experimental results on two datasets show that GAGS improves visual grounding accuracy by an average of 10.9% and semantic segmentation accuracy by an average of 7.0%, with an inference speed 2× faster than baseline methods.
Revisiting Cross-Architecture Distillation: Adaptive Dual-Teacher Transfer for Lightweight Video Models
PDF ↗Vision Transformers (ViTs) have achieved strong performance in video action recognition, but their high computational cost limits their practicality. Lightweight CNNs are more efficient but suffer from accuracy gaps. Cross-Architecture Knowledge Distillation (CAKD) addresses this by transferring knowledge from ViTs to CNNs, yet existing methods often struggle with architectural mismatch and overlook the value of stronger homogeneous CNN teachers. To tackle these challenges, we propose a Dual-Teacher Knowledge Distillation framework that leverages both a heterogeneous ViT teacher and a homogeneous CNN teacher to collaboratively guide a lightweight CNN student. We introduce two key components: (1) Discrepancy-Aware Teacher Weighting, which dynamically fuses the predictions from ViT and CNN teachers by assigning adaptive weights based on teacher confidence and prediction discrepancy with the student, enabling more informative and effective supervision; and (2) a Structure Discrepancy-Aware Distillation strategy, where the student learns the residual features between ViT and CNN teachers via a lightweight auxiliary branch, focusing on transferable architectural differences without mimicking all of ViT’s high-dimensional patterns. Extensive experiments on benchmarks including HMDB51, EPIC-KITCHENS-100, and Kinetics-400, demonstrate that our method consistently outperforms state-of-the-art distillation approaches, achieving notable performance improvements with a maximum accuracy gain of 5.95% on HMDB51.
3D Human Pose Estimation (3D HPE) is vital in various applications, from person re-identification and action recognition to virtual reality. However, the reliance on annotated 3D data collected in controlled environments poses challenges for generalization to diverse in-the-wild scenarios. Existing domain adaptation (DA) paradigms like general DA and source-free DA for 3D HPE overlook the issues of non-stationary target pose datasets. To address these challenges, we propose a novel task named lifelong domain adaptive 3D HPE. To our knowledge, we are the first to introduce the lifelong domain adaptation to the 3D HPE task. In this lifelong DA setting, the pose estimator is pretrained on the source domain and subsequently adapted to distinct target domains. Moreover, during adaptation to the current target domain, the pose estimator cannot access the source and all the previous target domains. The lifelong DA for 3D HPE involves overcoming challenges in adapting to current domain poses and preserving knowledge from previous domains, particularly combating catastrophic forgetting. We present an innovative Generative Adversarial Network (GAN) framework, which incorporates 3D pose generators, a 2D pose discriminator, and a 3D pose estimator. This framework effectively mitigates domain shifts and aligns original and augmented poses. Moreover, we construct a novel 3D pose generator paradigm, integrating pose-aware, temporal-aware, and domain-aware knowledge to enhance the current domain's adaptation and alleviate catastrophic forgetting on previous domains. Our method demonstrates superior performance through extensive experiments on diverse domain adaptive 3D HPE datasets.
During the video encoding process, the original spatial domain signal is first transformed into the frequency domain, followed by quantization and compression. As a result, the quality degradation in compressed videos primarily stems from distortions in the frequency domain information. However, existing video enhancement methods typically directly fuse information from adjacent frames in the spatial domain, making it difficult for models to effectively compensate for frequency domain distortions, which leads to suboptimal detail restoration. To address this issue, we propose a Hierarchical Frequency-Guided Alignment Transformer. Additionally, by analyzing the characteristics of the frequency domain, we find that different frequency bands exhibit both correlations and a certain degree of independence. Based on this, we introduce a Frequency-Aware Transformer module that employs a combination of independent and mixed processing to optimize information exchange across different frequency domains, effectively mitigating cross-interference from irrelevant information. Experimental results demonstrate that, compared to existing methods, our approach achieves state-of-the-art performance in objective metrics (PSNR/SSIM), perceptual quality (LPIPS), and subjective visual effects, while reducing model complexity.
Text-to-Video (T2V) generation has advanced greatly, yet maintaining consistency remains challenging, especially for tuning-free long video generation. We attribute the consistency problem to cumulative deviations for long video generation at three levels: the random noise lacking correlation results initial deviation between frames; discrepancy in semantic feature tokens between denoising network blocks gradually accumulates as the frame count grows, leading to greater deviations; attention mechanisms struggle to capture global relationships across distant frames in long videos. To address these, we propose FreeMem, a tuning-free framework leveraging hierarchical memory update and injection: the noise memory stabilizes consistency by manipulating low and high frequency components in the initial noise space; the token memory combats inconsistency through adaptive fusion of historical and current semantic feature tokens between denoising network blocks; and the attention memory establishes persistent cache to model long-range relationships within self attention layers. Evaluated on VBench, FreeMem improves subject and background consistency matrics across various methods, offering a practical solution for low-cost, high-consistency long video generation.