Shuo Ye

Minghui Jia, Qichao Zhang, Ali Luo et al.

Due to the limited generalization and interpretability of deep learning classifiers, The final vetting of rare celestial object candidates still relies on expert visual inspection--a manually intensive process. In this process, astronomers leverage specialized tools to analyze spectra and construct reliable catalogs. However, this practice has become the primary bottleneck, as it is fundamentally incapable of scaling with the data deluge from modern spectroscopic surveys. To bridge this gap, we propose Spec-o3, a tool-augmented vision-language agent that performs astronomer-aligned spectral inspection via interleaved multimodal chain-of-thought reasoning. Spec-o3 is trained with a two-stage post-training recipe: cold-start supervised fine-tuning on expert inspection trajectories followed by outcome-based reinforcement learning on rare-type verification tasks. Evaluated on five rare-object identification tasks from LAMOST, Spec-o3 establishes a new State-of-the-Art, boosting the macro-F1 score from 28.3 to 76.5 with a 7B parameter base model and outperforming both proprietary VLMs and specialized deep models. Crucially, the agent demonstrates strong generalization to unseen inspection tasks across survey shifts (from LAMOST to SDSS/DESI). Expert evaluations confirm that its reasoning traces are coherent and physically consistent, supporting transparent and trustworthy decision-making. Code, data, and models are available at https://github.com/Maxwell-Jia/spec-o3.

Yu Wang, Shuo Ye, Shujian Yu et al.

Fine-grained visual categorization (FGVC) aims to discriminate similar subcategories, whose main challenge is the large intraclass diversities and subtle inter-class differences. Existing FGVC methods usually select discriminant regions found by a trained model, which is prone to neglect other potential discriminant information. On the other hand, the massive interactions between the sequence of image patches in ViT make the resulting class-token contain lots of redundant information, which may also impacts FGVC performance. In this paper, we present a novel approach for FGVC, which can simultaneously make use of partial yet sufficient discriminative information in environmental cues and also compress the redundant information in class-token with respect to the target. Specifically, our model calculates the ratio of high-weight regions in a batch, adaptively adjusts the masking threshold and achieves moderate extraction of background information in the input space. Moreover, we also use the Information Bottleneck~(IB) approach to guide our network to learn a minimum sufficient representations in the feature space. Experimental results on three widely-used benchmark datasets verify that our approach can achieve outperforming performance than other state-of-the-art approaches and baseline models.

Xiaolong Tang, Shuo Ye, Yufeng Shi et al.

Filter pruning has gained widespread adoption for the purpose of compressing and speeding up convolutional neural networks (CNNs). However, existing approaches are still far from practical applications due to biased filter selection and heavy computation cost. This paper introduces a new filter pruning method that selects filters in an interpretable, multi-perspective, and lightweight manner. Specifically, we evaluate the contributions of filters from both individual and overall perspectives. For the amount of information contained in each filter, a new metric called information capacity is proposed. Inspired by the information theory, we utilize the interpretable entropy to measure the information capacity, and develop a feature-guided approximation process. For correlations among filters, another metric called information independence is designed. Since the aforementioned metrics are evaluated in a simple but effective way, we can identify and prune the least important filters with less computation cost. We conduct comprehensive experiments on benchmark datasets employing various widely-used CNN architectures to evaluate the performance of our method. For instance, on ILSVRC-2012, our method outperforms state-of-the-art methods by reducing FLOPs by 77.4% and parameters by 69.3% for ResNet-50 with only a minor decrease in accuracy of 2.64%.

Shuo Ye, Shujian Yu, Wenjin Hou et al.

Fine-grained visual categorization (FGVC) is a challenging task due to similar visual appearances between various species. Previous studies always implicitly assume that the training and test data have the same underlying distributions, and that features extracted by modern backbone architectures remain discriminative and generalize well to unseen test data. However, we empirically justify that these conditions are not always true on benchmark datasets. To this end, we combine the merits of invariant risk minimization (IRM) and information bottleneck (IB) principle to learn invariant and minimum sufficient (IMS) representations for FGVC, such that the overall model can always discover the most succinct and consistent fine-grained features. We apply the matrix-based R{é}nyi's $α$-order entropy to simplify and stabilize the training of IB; we also design a ``soft" environment partition scheme to make IRM applicable to FGVC task. To the best of our knowledge, we are the first to address the problem of FGVC from a generalization perspective and develop a new information-theoretic solution accordingly. Extensive experiments demonstrate the consistent performance gain offered by our IMS.

Tianxu Wu, Shuo Ye, Shuhuang Chen et al.

The challenge in fine-grained visual categorization lies in how to explore the subtle differences between different subclasses and achieve accurate discrimination. Previous research has relied on large-scale annotated data and pre-trained deep models to achieve the objective. However, when only a limited amount of samples is available, similar methods may become less effective. Diffusion models have been widely adopted in data augmentation due to their outstanding diversity in data generation. However, the high level of detail required for fine-grained images makes it challenging for existing methods to be directly employed. To address this issue, we propose a novel approach termed the detail reinforcement diffusion model~(DRDM), which leverages the rich knowledge of large models for fine-grained data augmentation and comprises two key components including discriminative semantic recombination (DSR) and spatial knowledge reference~(SKR). Specifically, DSR is designed to extract implicit similarity relationships from the labels and reconstruct the semantic mapping between labels and instances, which enables better discrimination of subtle differences between different subclasses. Furthermore, we introduce the SKR module, which incorporates the distributions of different datasets as references in the feature space. This allows the SKR to aggregate the high-dimensional distribution of subclass features in few-shot FGVC tasks, thus expanding the decision boundary. Through these two critical components, we effectively utilize the knowledge from large models to address the issue of data scarcity, resulting in improved performance for fine-grained visual recognition tasks. Extensive experiments demonstrate the consistent performance gain offered by our DRDM.

Shuo Ye, Shiming Chen, Ruxin Wang et al.

Data is the foundation for the development of computer vision, and the establishment of datasets plays an important role in advancing the techniques of fine-grained visual categorization~(FGVC). In the existing FGVC datasets used in computer vision, it is generally assumed that each collected instance has fixed characteristics and the distribution of different categories is relatively balanced. In contrast, the real world scenario reveals the fact that the characteristics of instances tend to vary with time and exhibit a long-tailed distribution. Hence, the collected datasets may mislead the optimization of the fine-grained classifiers, resulting in unpleasant performance in real applications. Starting from the real-world conditions and to promote the practical progress of fine-grained visual categorization, we present a Concept Drift and Long-Tailed Distribution dataset. Specifically, the dataset is collected by gathering 11195 images of 250 instances in different species for 47 consecutive months in their natural contexts. The collection process involves dozens of crowd workers for photographing and domain experts for labeling. Meanwhile, we propose a feature recombination framework to address the learning challenges associated with CDLT. Experimental results validate the efficacy of our method while also highlighting the limitations of popular large vision-language models (e.g., CLIP) in the context of long-tailed distributions. This emphasizes the significance of CDLT as a benchmark for investigating these challenges.

Yiyu Liu, Shuo Ye, Chao Hao et al.

Video Camouflaged Object Detection (VCOD) is currently constrained by the scarcity of challenging benchmarks and the limited robustness of models against erratic motion dynamics. Existing methods often struggle with Motion-Induced Appearance Instability and Temporal Feature Misalignment caused by complex motion scenarios. To address the data bottleneck, we present YUV20K, a pixel-level annoated complexity-driven VCOD benchmark. Comprising 24,295 annotated frames across 91 scenes and 47 kinds of species, it specifically targets challenging scenarios like large-displacement motion, camera motion and other 4 types scenarios. On the methodological front, we propose a novel framework featuring two key modules: Motion Feature Stabilization (MFS) and Trajectory-Aware Alignment (TAA). The MFS module utilizes frame-agnostic Semantic Basis Primitives to stablize features, while the TAA module leverages trajectory-guided deformable sampling to ensure precise temporal alignment. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art competitors on existing datasets and establishes a new baseline on the challenging YUV20K. Notably, our framework exhibits superior cross-domain generalization and robustness when confronting complex spatiotemporal scenarios. Our code and dataset will be available at https://github.com/K1NSA/YUV20K

Wenji Wu, Shuo Ye, Yiyu Liu et al.

Underwater Camouflaged Object Detection (UCOD) is a challenging task due to the extreme visual similarity between targets and backgrounds across varying marine depths. Existing methods often struggle with topological fragmentation of slender creatures in the deep sea and the subtle feature extraction of transparent organisms. In this paper, we propose DeepTopo-Net, a novel framework that integrates topology-aware modeling with frequency-decoupled perception. To address physical degradation, we design the Water-Conditioned Adaptive Perceptor (WCAP), which employs Riemannian metric tensors to dynamically deform convolutional sampling fields. Furthermore, the Abyssal-Topology Refinement Module (ATRM) is developed to maintain the structural connectivity of spindly targets through skeletal priors. Specifically, we first introduce GBU-UCOD, the first high-resolution (2K) benchmark tailored for marine vertical zonation, filling the data gap for hadal and abyssal zones. Extensive experiments on MAS3K, RMAS, and our proposed GBU-UCOD datasets demonstrate that DeepTopo-Net achieves state-of-the-art performance, particularly in preserving the morphological integrity of complex underwater patterns. The datasets and codes will be released at https://github.com/Wuwenji18/GBU-UCOD.

Jiayu Zhang, Shuo Ye, Qilang Ye et al.

Recent Audio-Visual Question Answering (AVQA) methods have advanced significantly. However, most AVQA methods lack effective mechanisms for handling missing modalities, suffering from severe performance degradation in real-world scenarios with data interruptions. Furthermore, prevailing methods for handling missing modalities predominantly rely on generative imputation to synthesize missing features. While partially effective, these methods tend to capture inter-modal commonalities but struggle to acquire unique, modality-specific knowledge within the missing data, leading to hallucinations and compromised reasoning accuracy. To tackle these challenges, we propose R$^{2}$ScP, a novel framework that shifts the paradigm of missing modality handling from traditional generative imputation to retrieval-based recovery. Specifically, we leverage cross-modal retrieval via unified semantic embeddings to acquire missing domain-specific knowledge. To maximize semantic restoration, we introduce a context-aware adaptive purification mechanism that eliminates latent semantic noise within the retrieved data. Additionally, we employ a two-stage training strategy to explicitly model the semantic relationships between knowledge from different sources. Extensive experiments demonstrate that R$^{2}$ScP significantly improves AVQA and enhances robustness in modal-incomplete scenarios.

Chao Hao, Jun Xu, Ji Du et al.

Language-guided segmentation transcends the scope limitations of traditional semantic segmentation, enabling models to segment arbitrary target regions based on natural language instructions. Existing approaches typically adopt a two-stage framework: employing Multimodal Large Language Models (MLLMs) to interpret instructions and generate visual prompts, followed by foundational segmentation models (e.g., SAM) to produce masks. However, due to the limited spatial grounding capabilities of off-the-shelf MLLMs, these methods often rely on extensive training on large-scale datasets to achieve satisfactory accuracy. While recent advances have introduced reasoning mechanisms to improve performance, they predominantly operate within the textual domain, performing chain-of-thought reasoning solely based on abstract text representations without direct visual feedback. In this paper, we propose Seg-Agent, a completely training-free framework that pioneers Explicit Multimodal Chain-of-Reasoning. Unlike prior text-only reasoning, our approach constructs an interactive visual reasoning loop comprising three stages: generation, selection, and refinement. Specifically, we leverage Set-of-Mark (SoM) visual prompting to render candidate regions directly onto the image, allowing the MLLM to ``see'' and iteratively reason about spatial relationships in the visual domain rather than just the textual one. This explicit multimodal interaction enables Seg-Agent to achieve performance comparable to state-of-the-art training-based methods without any parameter updates. Furthermore, to comprehensively evaluate generalization across diverse scenarios, we introduce Various-LangSeg, a novel benchmark covering explicit semantic, generic object, and reasoning-guided segmentation tasks. Extensive experiments demonstrate the effectiveness and robustness of our method.

Zhimeng Xin, Tianxu Wu, Shiming Chen et al.

Camouflaged object detection (COD) primarily relies on semantic or instance segmentation methods. While these methods have made significant advancements in identifying the contours of camouflaged objects, they may be inefficient or cost-effective for tasks that only require the specific location of the object. Object detection algorithms offer an optimized solution for Realistic Camouflaged Object Detection (RCOD) in such cases. However, detecting camouflaged objects remains a formidable challenge due to the high degree of similarity between the features of the objects and their backgrounds. Unlike segmentation methods that perform pixel-wise comparisons to differentiate between foreground and background, object detectors omit this analysis, further aggravating the challenge. To solve this problem, we propose a camouflage-aware feature refinement (CAFR) strategy. Since camouflaged objects are not rare categories, CAFR fully utilizes a clear perception of the current object within the prior knowledge of large models to assist detectors in deeply understanding the distinctions between background and foreground. Specifically, in CAFR, we introduce the Adaptive Gradient Propagation (AGP) module that fine-tunes all feature extractor layers in large detection models to fully refine class-specific features from camouflaged contexts. We then design the Sparse Feature Refinement (SFR) module that optimizes the transformer-based feature extractor to focus primarily on capturing class-specific features in camouflaged scenarios. To facilitate the assessment of RCOD tasks, we manually annotate the labels required for detection on three existing segmentation COD datasets, creating a new benchmark for RCOD tasks. Code and datasets are available at: https://github.com/zhimengXin/RCOD.

Zewen Li, Shuo Ye, Zitong Yu et al.

Industrial anomaly detection based on RGB-3D multimodal data has emerged as a mainstream paradigm for intelligent quality inspection. However, existing unsupervised methods suffer from two critical limitations: ambiguous cross-modal alignment caused by the lack of high-level semantic guidance and insufficient geometric modeling for RGB-to-3D feature mapping. To address these issues, we propose a unified multimodal industrial anomaly detection framework guided by text semantics. The framework consists of two core modules: a Geometry-Aware Cross-Modal Mapper to preserve geometric structure during modality conversion, and an Object-Conditioned Textual Feature Adaptor to align multimodal features with semantic priors. Furthermore, we establish a unified learning paradigm for multimodal industrial anomaly detection, which breaks the one-model-one-class constraint and enables accurate anomaly detection across diverse classes using a single model. Extensive experiments on the MVTec 3D-AD and Eyecandies datasets demonstrate that our method achieves state-of-the-art performance in classification and localization under unsupervised settings.

Dongliang Zhu, Zhiyi Niu, Bo Zhao et al.

Subtle visual signals, although difficult to perceive with the naked eye, contain important information that can reveal hidden patterns in visual data. These signals play a key role in many applications, including biometric security, multimedia forensics, medical diagnosis, industrial inspection, and affective computing. With the rapid development of computer vision and representation learning techniques, detecting and interpreting such subtle signals has become an emerging research direction. However, existing studies often focus on specific tasks or modalities, and models still face challenges in robustness, representation ability, and generalization when handling subtle and weak signals in real-world environments. To promote research in this area, we organize the Subtle visual Challenge, which aims to learn robust representations for subtle visual signals. The challenge includes two tasks: cross-domain multimodal deception detection and remote photoplethysmography (rPPG) estimation. We hope that this challenge will encourage the development of more robust and generalizable models for subtle visual understanding, and further advance research in computer vision and multimodal learning. A total of 22 teams submitted their final results to this workshop competition, and the corresponding baseline models have been released on the \href{https://sites.google.com/view/svc-cvpr26}{MMDD2026 platform}\footnote{https://sites.google.com/view/svc-cvpr26}

Peipei Yuan, Zijing Xie, Shuo Ye et al.

Generative artificial intelligence holds significant potential for abuse, and generative image detection has become a key focus of research. However, existing methods primarily focused on detecting a specific generative model and emphasizing the localization of synthetic regions, while neglecting the interference caused by image size and style on model learning. Our goal is to reach a fundamental conclusion: Is the image real or generated? To this end, we propose a diffusion model-based generative image detection framework termed Hierarchical Retrospection Refinement~(HRR). It designs a multi-scale style retrospection module that encourages the model to generate detailed and realistic multi-scale representations, while alleviating the learning biases introduced by dataset styles and generative models. Additionally, based on the principle of correntropy sparse additive machine, a feature refinement module is designed to reduce the impact of redundant features on learning and capture the intrinsic structure and patterns of the data, thereby improving the model's generalization ability. Extensive experiments demonstrate the HRR framework consistently delivers significant performance improvements, outperforming state-of-the-art methods in generated image detection task.

Jiayu Zhang, Qilang Ye, Shuo Ye et al.

Audio-Visual Question Answering (AVQA) requires models to effectively utilize both visual and auditory modalities to answer complex and diverse questions about audio-visual scenes. However, existing methods lack sufficient flexibility and dynamic adaptability in temporal sampling and modality preference awareness, making it difficult to focus on key information based on the question. This limits their reasoning capability in complex scenarios. To address these challenges, we propose a novel framework named AV-Master. It enhances the model's ability to extract key information from complex audio-visual scenes with substantial redundant content by dynamically modeling both temporal and modality dimensions. In the temporal dimension, we introduce a dynamic adaptive focus sampling mechanism that progressively focuses on audio-visual segments most relevant to the question, effectively mitigating redundancy and segment fragmentation in traditional sampling methods. In the modality dimension, we propose a preference-aware strategy that models each modality's contribution independently, enabling selective activation of critical features. Furthermore, we introduce a dual-path contrastive loss to reinforce consistency and complementarity across temporal and modality dimensions, guiding the model to learn question-specific cross-modal collaborative representations. Experiments on four large-scale benchmarks show that AV-Master significantly outperforms existing methods, especially in complex reasoning tasks.