论文检索

Weijia Liu, Jiuxin Cao, Bo Miao, Zhiheng Fu, Xuelin Zhu, Jiawei Ge, Bo Liu, Mehwish Nasim, Ajmal Mian

Current text-driven Video Moment Retrieval (VMR) methods encode all video clips, including irrelevant ones, disrupting multimodal alignment and hindering optimization. To this end, we propose a denoise-then-retrieve paradigm that explicitly filters text-irrelevant clips from videos and then retrieves the target moment using purified multimodal representations. Following this paradigm, we introduce the Denoise-then-Retrieve Network (DRNet), comprising Text-Conditioned Denoising (TCD) and Text-Reconstruction Feedback (TRF) modules. TCD integrates cross-attention and structured state space blocks to dynamically identify noisy clips and produce a noise mask to purify multimodal video representations. TRF further distills a single query embedding from purified video representations and aligns it with the text embedding, serving as auxiliary supervision for denoising during training. Finally, we perform conditional retrieval using text embeddings on purified video representations for accurate VMR. Experiments on Charades-STA and QVHighlights demonstrate that our approach surpasses state-of-the-art methods on all metrics. Furthermore, our denoise-then-retrieve paradigm is adaptable and can be seamlessly integrated into advanced VMR models to boost performance.

Wang Liu, Puhong Duan, Xudong Kang, Shutao Li

Segmenting ultra-high-resolution (UHR) images poses a significant challenge due to constraints on GPU memory, leading to a trade-off between detailed local information and a comprehensive contextual understanding. Current UHR methods often employ a multi-branch encoder to handle local and contextual information, which can be memory-intensive. To address the need for both high accuracy and low memory usage in processing UHR images, we introduce a memory-efficient semantic segmentation approach by squeezing context information into local patches (SCPSeg). Our method integrates the processing of local and contextual information within a single-branch encoder. Specifically, we introduce a context squeezing module (CSM) designed to compress global context details into local patches, enabling segmentation networks to perceive broader image contexts. Additionally, we propose a super-resolution guided local feature alignment (LFA) technique to improve segmentation precision by aligning local feature relationships. This approach calculates similarities within sliding windows, avoiding heavy computational costs during the training phase. We evaluate the effectiveness of our proposed method on four widely used UHR segmentation benchmarks. Experimental results demonstrate that our approach enhances UHR segmentation accuracy without incurring additional memory overhead during the inference stage. The code is available at https://github.com/StuLiu/SCPSeg.

Shuaiyu Liu, Qiyao Shen, Yunxi Wang, Yazhou Ren, Guoyin Wang

Facial Expression Recognition (FER) is a fundamental problem in computer vision. Despite recent advances, significant challenges remain. Current methods primarily focus on extracting visual representations while overlooking other valuable information. To address this limitation, we propose a novel method called Component Separation and Granular-ball Space Bootstrap Fusion (CS-GBSBF), which leverages granular balls to transform visual images to spatial graphs, thereby enlarging the spatial information embedded in images. Our method separates the face into different components and utilizes the spatial information to bootstrap the fusion. More specifically, CS-GBSBF mainly consists of three crucial networks: Represent Extraction Network (REN), Represent Separation Network (RSN) and Represent Fusion Network (RFN). First, granular balls are used to represent expression images as graphs, which are fed into REN along with images. Then, RSN separates basic visual/spatial representations extracted from REN into a set of component visual/spatial representations. Next, RFN utilizes spatial representations to bootstrap component visual integration. A significant challenge in two-stream models is feature alignment, for which we have developed Attention Guidance Module (AGM) and Bootstrap Alignment Loss (L_BA) in REN and RFN, respectively. Results of experiment on eight databases show that CS-GBSBF consistently achieves higher recognition accuracy than several state-of-the-art methods. The code is available at https://github.com/Lsy235/CS-GBSBF.

Ming-Hui Liu, Harry Cheng, Tianyi Wang, Xin Luo, Xin-Shun Xu

Deepfake detection models often struggle with generalization to unseen datasets, manifesting as misclassifying real instances as fake in target domains. This is primarily due to an overreliance on forgery artifacts and a limited understanding of real faces. To address this challenge, we propose a novel approach RealID to enhance generalization by learning a comprehensive concept of real faces while assessing the probabilities of belonging to the real and fake classes independently. RealID comprises two key modules: the Real Concept Capture Module (RealC^2) and the Independent Dual-Decision Classifier (IDC). With the assistance of a Multi-Real Memory, RealC^2 maintains various prototypes for real faces, allowing the model to capture a comprehensive concept of real class. Meanwhile, IDC redefines the classification strategy by making independent decisions based on the concept of the real class and the presence of forgery artifacts. Through the combined effect of the above modules, the influence of forgery-irrelevant patterns is alleviated, and extensive experiments on five widely used datasets demonstrate that RealID significantly outperforms existing state-of-the-art methods, achieving a 1.74% improvement in average accuracy.

Ke Liu, Jiwei Wei, Shiyuan He, Zeyu Ma, Chaoning Zhang, Ning Xie, Yang Yang

Generating high-fidelity talking heads that maintain stable head poses and achieve robust lip sync remains a significant challenge. Although methods based on 3D Gaussian Splatting (3DGS) offer a promising solution via point-based deformation, they suffer from inconsistent head dynamics and mismatched mouth movements due to unstable Gaussian initialization and incomplete speech features. To overcome these limitations, we introduce SyncGaussian, a 3DGS-based framework that ensures stable head poses, enhanced lip sync, and realistic appearances with real-time rendering. SyncGaussian employs a stable head Gaussian initialization strategy to mitigate head jitter by optimizing commonly used rough head pose parameters. To enhance lip sync, we propose a sync-enhanced encoder that leverages audio-to-text and audio-to-visual speech features. Guided by a tailored cosine similarity loss function, the encoder integrates discriminative speech features through a multi-level sync adaptation mechanism, enabling the learning of an adaptive speech feature space. Extensive experiments demonstrate that SyncGaussian outperforms state-of-the-art methods in image quality, dynamic motion, and lip sync, with the potential for real-time applications.

Jinyuan Liu, Qianjun Huang, Guanyao Wu, Di Wang, Zhiying Jiang, Long Ma, Risheng Liu, Xin Fan

Multi-exposure image fusion~(MEF) aims to integrate a set of low dynamic range images, producing a single image with a higher dynamic range than either one. Despite significant advancements, current MEF approaches still struggle to handle extremely over- or under-exposed conditions, resulting in unsatisfactory visual effects such as hallucinated details and distorted color tones. With this regard, we propose TextMEF, a prompt-driven fusion method enhanced by prompt learning, for multi-exposure image fusion. Specifically, we learn a set of prompts based on text-image similarity among negative and positive samples (over-exposed, under-exposed images, and well-exposed ones). These learned prompts are seamlessly integrated into the loss function, providing high-level guidance for constraining non-uniform exposure regions. Furthermore, we develop a attention Mamba module effectively translates over-/under- exposed regional features into exposure invariant space and ensure them to build efficient long-range dependency to high dynamic range image. Extensive experimental results on three publicly available benchmarks demonstrate that our TextMEF significantly outperforms state-of-the-art approaches in both visual inspection and objective analysis.

Biao Liu, Xu Liu, Lingling Li, Licheng Jiao, Fang Liu, Xinyu Sun, Youlin Huang

Visual grounding (VG) refers to detecting the specific objects in images based on linguistic expressions, and it has profound significance in the advanced interpretation of natural images. In remote sensing image interpretation, visual grounding is limited by characteristics such as the complex scenes and diverse object sizes. To solve this problem, we propose a novel remote sensing visual grounding (RSVG) framework, named language-guided hybrid representation learning Transformer (LGFormer). Specifically, we designed a multimodal dual-encoder Transformer structure called the adaptive multimodal feature fusion module. This structure innovatively integrates text and visual features as hybrid queries, enabling early-stage decoding queries to perceive the target position accurately. Then, the different modal information from the dual encoders is aggregated by hybrid queries to obtain the final object embedding for coordinate regression. Besides, a multi-scale cross-modal feature enhancement module (MSCM) is designed to enhance the self-representation of the extracted text and visual features and align them semantically. As for the hybrid queries, we use linguistic guidance to select visual features as the visual part and sentence-level features as the textual part. Finally, the LGFormer model we designed achieved the best results compared to existing models on the DIOR-RSVG and OPT-RSVG datasets.

Yihong Lin, Zhaoxin Fan, Xianjia Wu, Lingyu Xiong, Xiandong Li, Wenxiong Kang, Liang Peng, Songju Lei, Huang Xu

Speech-driven talking head generation is a critical yet challenging task with applications in augmented reality and virtual human modeling. While recent approaches using autoregressive and diffusion-based models have achieved notable progress, they often suffer from modality inconsistencies, particularly misalignment between audio and mesh, leading to reduced motion diversity and lip-sync accuracy. To address this, we propose GLDiTalker, a novel speech-driven 3D facial animation model based on a Graph Latent Diffusion Transformer. GLDiTalker resolves modality misalignment by diffusing signals within a quantized spatiotemporal latent space. It employs a two-stage training pipeline: the Graph-Enhanced Quantized Space Learning Stage ensures lip-sync accuracy, while the Space-Time Powered Latent Diffusion Stage enhances motion diversity. Together, these stages enable GLDiTalker to generate realistic, temporally stable 3D facial animations. Extensive evaluations on standard benchmarks demonstrate that GLDiTalker outperforms existing methods, achieving superior results in both lip-sync accuracy and motion diversity.

Shuyuan Lin, Mengtin Lo, Haosheng Chen, Yanjie Liang, Qiangqiang Wu

Two-view correspondence learning is a key task in computer vision, which aims to establish reliable matching relationships for applications such as camera pose estimation and 3D reconstruction. However, existing methods have limitations in local geometric modeling and cross-stage information optimization, which make it difficult to accurately capture the geometric constraints of matched pairs and thus reduce the robustness of the model. To address these challenges, we propose a Multi-Graph Contextual Attention Network (MGCA-Net), which consists of a Contextual Geometric Attention (CGA) module and a Cross-Stage Multi-Graph Consensus (CSMGC) module. Specifically, CGA dynamically integrates spatial position and feature information via an adaptive attention mechanism and enhances the capability to capture both local and global geometric relationships. Meanwhile, CSMGC establishes geometric consensus via a cross-stage sparse graph network, ensuring the consistency of geometric information across different stages. Experimental results on two representative YFCC100M and SUN3D datasets show that MGCA-Net significantly outperforms existing SOTA methods in the outlier rejection and camera pose estimation tasks. Source code is available at http://www.linshuyuan.com.

Zhechun Liang, Tao Huang, Fangfang Wu, Shiwen Xue, Zhenyu Wang, Weisheng Dong, Xin Li, Guangming Shi

Remote sensing composed image retrieval (RSCIR) is a new vision-language task that takes a composed query of an image and text, aiming to search for a target remote sensing image satisfying two conditions from intricate remote sensing imagery. However, the existing attribute-based benchmark Patterncom in RSCIR has significant flaws, including the lack of query text sentences and paired triplets, thus making it unable to evaluate the latest methods. To address this, we propose the Zero-Shot Query Text Generator (ZS-QTG) that can generate full query text sentences based on attributes, and then, by capitalizing on ZS-QTG, we develop the PatternCIR benchmark. PatternCIR rectifies Patterncom’s deficiencies and enables the evaluation of existing methods. Additionally, we explore zero-shot composed image retrieval methods that do not rely on massive pre-collected triplets for training. Existing methods use only the text during retrieval, performing poorly in RSCIR. To improve this, we propose Text-image Sequential Training of Composed Image Retrieval (TisCIR). TisCIR undergoes sequential training of multiple self-masking projection and fine-grained image attention modules, which endows it with the capacity to filter out conflicting information between the image and text, enhancing the retrieval by utilizing both modalities in harmony. TisCIR outperforms existing methods by 12.40% to 62.03% on PatternCIR, achieving state-of-the-art performance in RSCIR. The data and code are available here.

Jiawei Liang, Siyuan Liang, Jianjie Huang, Chenxi Si, Ming Zhang, Xiaochun Cao

The advancement of deep object detectors has greatly affected safety-critical fields like autonomous driving. However, physical adversarial camouflage poses a significant security risk by altering object textures to deceive detectors. Existing techniques struggle with variable physical environments, facing two main challenges: 1) inconsistent sampling point densities across distances hinder the gradient optimization from ensuring local continuity, and 2) updating texture gradients from multiple angles causes conflicts, reducing optimization stability and attack effectiveness. To address these issues, we propose a novel adversarial camouflage framework based on gradient optimization. First, we introduce a gradient calibration strategy, which ensures consistent gradient updates across distances by propagating gradients from sparsely to unsampled texture points, thereby expanding the attack's effective range. Additionally, we develop a gradient decorrelation method, which prioritizes and orthogonalizes gradients based on loss values, enhancing stability and effectiveness in multi-angle optimization by eliminating redundant or conflicting updates. Extensive experimental results on various detection models, angles, and distances show that our method significantly surpasses the state-of-the-art, with an average attack success rate (ASR) increase of 13.46\% across distances and 11.03\% across angles. Furthermore, experiments in real-world settings confirm the method's threat potential, highlighting the urgent need for more robust autopilot systems less prone to spoofing.

Zixuan Li, Lei Meng, Guoqing Chao, Wei Wu, Yimeng Yang, Xiaoshuo Yan, Zhuang Qi, Xiangxu Meng

Cross-modal alignment is an effective approach to improving visual classification. Existing studies typically enforce a one-step mapping that uses deep neural networks to project the visual features to mimic the distribution of textual features. However, they typically face difficulties in finding such a projection due to the two modalities in both the distribution of class-wise samples and the range of their feature values. To address this issue, this paper proposes a novel Semantic-Space-Intervened Diffusive Alignment method, termed SeDA, models a semantic space as a bridge in the visual-to-textual projection, considering both types of features share the same class-level information in classification. More importantly, a bi-stage diffusion framework is developed to enable the progressive alignment between the two modalities. Specifically, SeDA first employs a Diffusion-Controlled Semantic Learner to model the semantic feature space of visual features by constraining the interactive features of the diffusion model and the category centers of visual features. In the later stage of SeDA, the Diffusion-Controlled Semantic Translator focuses on learning the distribution of textual features from the semantic space. Meanwhile, the Progressive Feature Interaction Network introduces stepwise feature interactions at each alignment step, progressively integrating textual information into mapped features. Experimental results show that SeDA achieves stronger cross-modal feature alignment, leading to superior performance over existing methods across multiple scenarios.

Yueying Li, Hanbin Zhao, Jiaqing Zhou, Guozhi Xu, Tianlei Hu, Gang Chen, Haobo Wang

Image super-resolution (ISR) is a classic and challenging problem in computer vision because of complex and unknown degradation patterns in the data collection process. Leveraging powerful generative priors, diffusion-based methods have recently established new state-of-the-art ISR performance, but their characteristics in the frequency domain are still underexplored. In this paper, we innovatively investigate their frequency-domain behaviors from a sampling timestep perspective. Experimentally, we find that current diffusion-based ISR algorithms exhibit insufficiency in different frequency components in distinct groups of timesteps during the sampling. To address this, we first propose a Timestep Division Controller that is able to adaptively divide the timesteps into groups based on the performance gradient across different components. Next, we design two dedicated modules --- the Amplitude and Phase Enhancement Module (APEM) and the High- and Low-Frequency Enhancement Module (HLEM), to regulate the information flow of distinct frequency-domain features. By adaptively enhancing specific frequency components at different stages of the sampling process, the two modules effectively compensate for the insufficient frequency-domain perception of diffusion-based ISR models. Extensive experiments on three benchmark datasets verify the superior ISR performance of our method, e.g., achieving an average 5.40% improvement on CLIP-IQA compared to the best diffusion-based ISR baseline.

Yongqian Li, Yong Luo, Xiantao Cai, Zheng He, Zhennan Meng, Nidong Wang, Yunlin Chen, Zhifei Li

We proposes a novel visual tokenizer by combining high-level semantic tokens and low-level pixel tokens to represent images, aiming to address the challenges of image-to-sequence conversion for Large Language Models (LLMs). Existing visual tokenizers, such as VQ-VAE and diffusion-based models, either struggle with token explosion as image resolution increases or fail to capture detailed structural information. Our method introduces a dual-token system: high-level semantic tokens capture the main content of the image, while low-level pixel tokens preserve structural details. By integrating these tokens in a hybrid architecture, we leverage a VQ-VAE branch to generate low-resolution guidance and a diffusion process to reconstruct high-resolution images with both semantic coherence and structural accuracy. This approach significantly reduces the number of required tokens and enhances image reconstruction quality, offering an efficient solution for tasks like image generation and understanding based on LLMs.

Yiming Li, Xiaoshan Yang, Bing-Kun Bao, Changsheng Xu

Due to the dynamic nature in videos, it is evident that perceiving and reasoning about temporal information are the key focus of Video Question Answering (VideoQA). In recent years, several methods have explored relationship-level temporal modeling with graph-structured video representation. Unfortunately, these methods heavily rely on the question text, thus making it challenging to perceive and reason about video content that is not explicitly mentioned in the question. To address the above challenge, we propose Graph Prompts-based VideoQA (GP-VQA), which adopts a video-based graph structure for enhanced video understanding. The proposed GP-VQA contains two stages, i.e., pre-training and prompt tuning. In pre-training, we define the pretext task that requires GP-VQA to reason about the randomly masked nodes or edges in the video graph, thus prompting GP-VQA to learn the reasoning ability with video-guided information. In prompt-tuning, we organize the textual question into question graph and implement message passing from video graph to question graph, therefore inheriting the video-based reasoning ability from video graph completion to VideoQA. Extensive experiments on various datasets have demonstrated the promising performance of GP-VQA.

Yao Li, Yong Zhou, Jiaqi Zhao, Wen-liang Du, Rui Yao, Bing Liu

Traditional unsupervised domain adaptation (UDA) struggles to extract rich semantics due to backbone limitations. Recent large-scale pre-trained visual-language models (VLMs) have shown strong zero-shot learning capabilities in UDA tasks. However, directly using VLMs results in a mixture of semantic and domain-specific information, complicating knowledge transfer. Complex scenes with subtle semantic differences are prone to misclassification, which in turn can result in the loss of features that are crucial for distinguishing between classes. To address these challenges, we propose a novel counterfactual knowledge maintenance UDA framework. Specifically, we employ counterfactual disentanglement to separate the representation of semantic information from domain features, thereby reducing domain bias. Furthermore, to clarify ambiguous visual information specific to classes, we maintain the discriminative knowledge of both visual and textual information. This approach synergistically leverages multimodal information to preserve modality-specific distinguishable features. We conducted extensive experimental evaluations on several public datasets to demonstrate the effectiveness of our method. The source code is available at https://github.com/LiYaolab/CMKUDA

Xulin Li, Yan Lu, Bin Liu, Jiaze Li, Qinhong Yang, Tao Gong, Qi Chu, Mang Ye, Nenghai Yu

In real applications, person re-identification (ReID) expects to retrieve the target person at any time, including both daytime and nighttime, ranging from short-term to long-term. However, existing ReID tasks and datasets cannot meet this requirement, as they are constrained by available time and only provide training and evaluation for specific scenarios. Therefore, we investigate a new task called Anytime Person Re-identification (AT-ReID), which aims to achieve effective retrieval in multiple scenarios based on variations in time. To address the AT-ReID problem, we collect the first large-scale dataset, AT-USTC, which contains 135k images of individuals wearing multiple clothes captured by RGB and IR cameras. Our data collection spans over an entire year and 270 volunteers were photographed on average 29.1 times across different dates or scenes, 4-15 times more than current datasets, providing conditions for follow-up investigations in AT-ReID. Further, to tackle the new challenge of multi-scenario retrieval, we propose a unified model named Uni-AT, which comprises a multi-scenario ReID (MS-ReID) framework for scenario-specific features learning, a Mixture-of-Attribute-Experts (MoAE) module to alleviate inter-scenario interference, and a Hierarchical Dynamic Weighting (HDW) strategy to ensure balanced training across all scenarios. Extensive experiments show that our model leads to satisfactory results and exhibits excellent generalization to all scenarios.

Xinyu Li, Qi Yao, Yuanda Wang

Garment sewing patterns are fundamental design elements that bridge the gap between design concepts and practical manufacturing. The generative modeling of sewing patterns is crucial for creating diversified garments. However, existing approaches are limited either by reliance on a single input modality or by suboptimal generation efficiency. In this work, we present GarmentDiffusion, a new generative model capable of producing centimeter-precise, vectorized 3D sewing patterns from multimodal inputs (text, image, and incomplete sewing pattern). Our method efficiently encodes 3D sewing pattern parameters into compact edge token representations, achieving a sequence length that is 10 times shorter than that of the autoregressive SewingGPT in DressCode. By employing a diffusion transformer, we simultaneously denoise all edge tokens along the temporal axis, while maintaining a constant number of denoising steps regardless of dataset-specific edge and panel statistics. With all combination of designs of our model, the sewing pattern generation speed is accelerated by 100 times compared to SewingGPT. We achieve new state-of-the-art results on DressCodeData, as well as on the largest sewing pattern dataset, namely GarmentCodeData. The project website is available at https://shenfu-research.github.io/Garment-Diffusion.

Xiangyue Li, Xiaoyang Wang, Zhibin Wan, Quan Zhang, Yupei Wu, Tao Deng, Mingjie Sun

This paper aims at generating anomalous images and their segmentation labels to address the lack of real-world anomaly samples and privacy issues. Departing from conventional approaches that use masks solely to guide the generation of anomaly images, we propose a dual-branch training strategy for the generative model. This strategy enables the simultaneous production of anomaly images and masks, with an alignment regularization loss that ensures the coherence between the generated images and their masks. During inference, only the image-generation branch is activated to produce synthetic samples for training the downstream segmentation model. Furthermore, we propose to integrate the well-trained generative model into the training of segmentation models, utilizing a generative feedback loss to refine the segmentation model's performance. Experiments show our method's IoU metrics exceed previous methods by 5.03%, 5.68% and 16.63% on Real-IAD (industrial), polyp (medical), and Floor Dirty (indoor) datasets. The code is publicly accessible at https://github.com/huan-yin/anomaly-alignment.

Weishi Li, Yong Peng, Mengyao Du, Fuhui Sun, Xiaoyan Wang, Li Shen

Personalized Federated Learning (PFL) meets each user’s personalized needs while still facing the high communication costs due to the large amount of data transmission and frequent communication. Decentralized PFL (DPFL) as an alternative discards the central server in PFL, which reduces the pressure of communication and the risk of server failure by using peer-to-peer communication.Nevertheless, DPFL still suffers from the significant communication pressure due to the transmission of a large number of model parameters, especially numerous nodes. To address the issues, we propose a novel personalized framework, DFedHP, in which each client utilizes a hypernetwork to generate the shared part of model parameters and train the personalized parameters separately. The number of parameters in a hypernetwork is much smaller than those in a typical local network, so hypernetwork aggregation reduces communication costs and the risk of privacy leakage. Furthermore, DFedHP can seamlessly integrate into existing DPFL algorithms as a plugin to boost their efficacy. At last, extensive experiments on various data heterogeneous environments demonstrate that DFedHP can reduce communication costs, accelerate convergence rate, and improve generalization performance compared with state-of-the-art (SOTA) baselines.