Showing 1–50 of 83 results for author: Pei, W

HISA: Efficient Hierarchical Indexing for Fine-Grained Sparse Attention

Authors: Yufei Xu, Fanxu Meng, Fan Jiang, Yuxuan Wang, Ruijie Zhou, Zhaohui Wang, Jiexi Wu, Zhixin Pan, Xiaojuan Tang, Wenjie Pei, Tongxuan Liu, Di Yin, Xing Sun, Muhan Zhang

Abstract: Token-level sparse attention mechanisms, exemplified by DeepSeek Sparse Attention (DSA), achieve fine-grained key selection by scoring every historical key for each query through a lightweight indexer, then computing attention only on the selected subset. While the downstream sparse attention itself scales favorably, the indexer must still scan the entire prefix for every query, introducing an per… ▽ More Token-level sparse attention mechanisms, exemplified by DeepSeek Sparse Attention (DSA), achieve fine-grained key selection by scoring every historical key for each query through a lightweight indexer, then computing attention only on the selected subset. While the downstream sparse attention itself scales favorably, the indexer must still scan the entire prefix for every query, introducing an per-layer bottleneck that grows prohibitively with context length. We propose HISA (Hierarchical Indexed Sparse Attention), a plug-and-play replacement for the indexer that rewrites the search path from a flat token scan into a two-stage hierarchical procedure: (1) a block-level coarse filtering stage that scores pooled block representations to discard irrelevant regions, followed by (2) a token-level refinement stage that applies the original indexer exclusively within the retained candidate blocks. HISA preserves the identical token-level top-sparse pattern consumed by the downstream Sparse MLA operator and requires no additional training. On kernel-level benchmarks, HISA achieves up to speedup at 64K context. On Needle-in-a-Haystack and LongBench, we directly replace the indexer in DeepSeek-V3.2 and GLM-5 with our HISA indexer, without any finetuning. HISA closely matches the original DSA in quality, while substantially outperforming block-sparse baselines. △ Less

Submitted 6 April, 2026; v1 submitted 30 March, 2026; originally announced March 2026.

Too Vivid to Be Real? Benchmarking and Calibrating Generative Color Fidelity

Authors: Zhengyao Fang, Zexi Jia, Yijia Zhong, Pengcheng Luo, Jinchao Zhang, Guangming Lu, Jun Yu, Wenjie Pei

Abstract: Recent advances in text-to-image (T2I) generation have greatly improved visual quality, yet producing images that appear visually authentic to real-world photography remains challenging. This is partly due to biases in existing evaluation paradigms: human ratings and preference-trained metrics often favor visually vivid images with exaggerated saturation and contrast, which make generations often… ▽ More Recent advances in text-to-image (T2I) generation have greatly improved visual quality, yet producing images that appear visually authentic to real-world photography remains challenging. This is partly due to biases in existing evaluation paradigms: human ratings and preference-trained metrics often favor visually vivid images with exaggerated saturation and contrast, which make generations often too vivid to be real even when prompted for realistic-style images. To address this issue, we present Color Fidelity Dataset (CFD) and Color Fidelity Metric (CFM) for objective evaluation of color fidelity in realistic-style generations. CFD contains over 1.3M real and synthetic images with ordered levels of color realism, while CFM employs a multimodal encoder to learn perceptual color fidelity. In addition, we propose a training-free Color Fidelity Refinement (CFR) that adaptively modulates spatial-temporal guidance scale in generation, thereby enhancing color authenticity. Together, CFD supports CFM for assessment, whose learned attention further guides CFR to refine T2I fidelity, forming a progressive framework for assessing and improving color fidelity in realistic-style T2I generation. The dataset and code are available at https://github.com/ZhengyaoFang/CFM. △ Less

Submitted 11 March, 2026; originally announced March 2026.

Comments: accepted by CVPR2026

Prune Redundancy, Preserve Essence: Vision Token Compression in VLMs via Synergistic Importance-Diversity

Authors: Zhengyao Fang, Pengyuan Lyu, Chengquan Zhang, Guangming Lu, Jun Yu, Wenjie Pei

Abstract: Vision-language models (VLMs) face significant computational inefficiencies caused by excessive generation of visual tokens. While prior work shows that a large fraction of visual tokens are redundant, existing compression methods struggle to balance importance preservation and information diversity. To address this, we propose PruneSID, a training-free Synergistic Importance-Diversity approach fe… ▽ More Vision-language models (VLMs) face significant computational inefficiencies caused by excessive generation of visual tokens. While prior work shows that a large fraction of visual tokens are redundant, existing compression methods struggle to balance importance preservation and information diversity. To address this, we propose PruneSID, a training-free Synergistic Importance-Diversity approach featuring a two-stage pipeline: (1) Principal Semantic Components Analysis (PSCA) for clustering tokens into semantically coherent groups, ensuring comprehensive concept coverage, and (2) Intra-group Non-Maximum Suppression (NMS) for pruning redundant tokens while preserving key representative tokens within each group. Additionally, PruneSID incorporates an information-aware dynamic compression ratio mechanism that optimizes token compression rates based on image complexity, enabling more effective average information preservation across diverse scenes. Extensive experiments demonstrate state-of-the-art performance, achieving 96.3% accuracy on LLaVA-1.5 with only 11.1% token retention, and 92.8% accuracy at extreme compression rates (5.6%) on LLaVA-NeXT, outperforming prior methods by 2.5% with 7.8 $\times$ faster prefilling speed compared to the original model. Our framework generalizes across diverse VLMs and both image and video modalities, showcasing strong cross-modal versatility. Code is available at https://github.com/ZhengyaoFang/PruneSID. △ Less

Submitted 11 March, 2026; v1 submitted 10 March, 2026; originally announced March 2026.

Comments: accepted by ICLR2026

DiffTrans: Differentiable Geometry-Materials Decomposition for Reconstructing Transparent Objects

Authors: Changpu Li, Shuang Wu, Songlin Tang, Guangming Lu, Jun Yu, Wenjie Pei

Abstract: Reconstructing transparent objects from a set of multi-view images is a challenging task due to the complicated nature and indeterminate behavior of light propagation. Typical methods are primarily tailored to specific scenarios, such as objects following a uniform topology, exhibiting ideal transparency and surface specular reflections, or with only surface materials, which substantially constrai… ▽ More Reconstructing transparent objects from a set of multi-view images is a challenging task due to the complicated nature and indeterminate behavior of light propagation. Typical methods are primarily tailored to specific scenarios, such as objects following a uniform topology, exhibiting ideal transparency and surface specular reflections, or with only surface materials, which substantially constrains their practical applicability in real-world settings. In this work, we propose a differentiable rendering framework for transparent objects, dubbed DiffTrans, which allows for efficient decomposition and reconstruction of the geometry and materials of transparent objects, thereby reconstructing transparent objects accurately in intricate scenes with diverse topology and complex texture. Specifically, we first utilize FlexiCubes with dilation and smoothness regularization as the iso-surface representation to reconstruct an initial geometry efficiently from the multi-view object silhouette. Meanwhile, we employ the environment light radiance field to recover the environment of the scene. Then we devise a recursive differentiable ray tracer to further optimize the geometry, index of refraction and absorption rate simultaneously in a unified and end-to-end manner, leading to high-quality reconstruction of transparent objects in intricate scenes. A prominent advantage of the designed ray tracer is that it can be implemented in CUDA, enabling a significantly reduced computational cost. Extensive experiments on multiple benchmarks demonstrate the superior reconstruction performance of our DiffTrans compared with other methods, especially in intricate scenes involving transparent objects with diverse topology and complex texture. The code is available at https://github.com/lcp29/DiffTrans. △ Less

Submitted 27 February, 2026; originally announced March 2026.

Breaking the Blocks: Continuous Low-Rank Decomposed Scaling for Unified LLM Quantization and Adaptation

Authors: Pingzhi Tang, Ruijie Zhou, Fanxu Meng, Wenjie Pei, Muhan Zhang

Abstract: Current quantization methods for LLMs predominantly rely on block-wise structures to maintain efficiency, often at the cost of representational flexibility. In this work, we demonstrate that element-wise quantization can be made as efficient as block-wise scaling while providing strictly superior expressive power by modeling the scaling manifold as continuous low-rank matrices ($S = BA$). We propo… ▽ More Current quantization methods for LLMs predominantly rely on block-wise structures to maintain efficiency, often at the cost of representational flexibility. In this work, we demonstrate that element-wise quantization can be made as efficient as block-wise scaling while providing strictly superior expressive power by modeling the scaling manifold as continuous low-rank matrices ($S = BA$). We propose Low-Rank Decomposed Scaling (LoRDS), a unified framework that rethinks quantization granularity through this low-rank decomposition. By "breaking the blocks" of spatial constraints, LoRDS establishes a seamless efficiency lifecycle: it provides high-fidelity PTQ initialization refined via iterative optimization, enables joint QAT of weights and scaling factors, and facilitates high-rank multiplicative PEFT adaptation. Unlike additive PEFT approaches such as QLoRA, LoRDS enables high-rank weight updates within a low-rank budget while incurring no additional inference overhead. Supported by highly optimized Triton kernels, LoRDS consistently outperforms state-of-the-art baselines across various model families in both quantization and downstream fine-tuning tasks. Notably, on Llama3-8B, our method achieves up to a 27.0% accuracy improvement at 3 bits over NormalFloat quantization and delivers a 1.5x inference speedup on NVIDIA RTX 4090 while enhancing PEFT performance by 9.6% on downstream tasks over 4bit QLoRA, offering a robust and integrated solution for unified compression and adaptation of LLMs. △ Less

Submitted 30 January, 2026; originally announced January 2026.

Evo-TFS: Evolutionary Time-Frequency Domain-Based Synthetic Minority Oversampling Approach to Imbalanced Time Series Classification

Authors: Wenbin Pei, Ruohao Dai, Bing Xue, Mengjie Zhang, Qiang Zhang, Yiu-Ming Cheung

Abstract: Time series classification is a fundamental machine learning task with broad real-world applications. Although many deep learning methods have proven effective in learning time-series data for classification, they were originally developed under the assumption of balanced data distributions. Once data distribution is uneven, these methods tend to ignore the minority class that is typically of high… ▽ More Time series classification is a fundamental machine learning task with broad real-world applications. Although many deep learning methods have proven effective in learning time-series data for classification, they were originally developed under the assumption of balanced data distributions. Once data distribution is uneven, these methods tend to ignore the minority class that is typically of higher practical significance. Oversampling methods have been designed to address this by generating minority-class samples, but their reliance on linear interpolation often hampers the preservation of temporal dynamics and the generation of diverse samples. Therefore, in this paper, we propose Evo-TFS, a novel evolutionary oversampling method that integrates both time- and frequency-domain characteristics. In Evo-TFS, strongly typed genetic programming is employed to evolve diverse, high-quality time series, guided by a fitness function that incorporates both time-domain and frequency-domain characteristics. Experiments conducted on imbalanced time series datasets demonstrate that Evo-TFS outperforms existing oversampling methods, significantly enhancing the performance of time-domain and frequency-domain classifiers. △ Less

Submitted 3 January, 2026; originally announced January 2026.

HWL-HIN: A Hypergraph-Level Hypergraph Isomorphism Network as Powerful as the Hypergraph Weisfeiler-Lehman Test with Application to Higher-Order Network Robustness

Authors: Chengyu Tian, Wenbin Pei

Abstract: Robustness in complex systems is of significant engineering and economic importance. However, conventional attack-based a posteriori robustness assessments incur prohibitive computational overhead. Recently, deep learning methods, such as Convolutional Neural Networks (CNNs) and Graph Neural Networks (GNNs), have been widely employed as surrogates for rapid robustness prediction. Nevertheless, the… ▽ More Robustness in complex systems is of significant engineering and economic importance. However, conventional attack-based a posteriori robustness assessments incur prohibitive computational overhead. Recently, deep learning methods, such as Convolutional Neural Networks (CNNs) and Graph Neural Networks (GNNs), have been widely employed as surrogates for rapid robustness prediction. Nevertheless, these methods neglect the complex higher-order correlations prevalent in real-world systems, which are naturally modeled as hypergraphs. Although Hypergraph Neural Networks (HGNNs) have been widely adopted for hypergraph learning, their topological expressive power has not yet reached the theoretical upper bound. To address this limitation, inspired by Graph Isomorphism Networks, this paper proposes a hypergraph-level Hypergraph Isomorphism Network framework. Theoretically, this approach is proven to possess an expressive power strictly equivalent to the Hypergraph Weisfeiler-Lehman test and is applied to predict hypergraph robustness. Experimental results demonstrate that while maintaining superior efficiency in training and prediction, the proposed method not only outperforms existing graph-based models but also significantly surpasses conventional HGNNs in tasks that prioritize topological structure representation. △ Less

Submitted 26 December, 2025; originally announced December 2025.

Exploring the Effectiveness of Google Play Store's Privacy Transparency Channels

Authors: Anhao Xiang, Weiping Pei, Chuan Yue

Abstract: With the requirements and emphases on privacy transparency placed by regulations such as GDPR and CCPA, the Google Play Store requires Android developers to more responsibly communicate their apps' privacy practices to potential users by providing the proper information via the data safety, privacy policy, and permission manifest privacy transparency channels. However, it is unclear how effective… ▽ More With the requirements and emphases on privacy transparency placed by regulations such as GDPR and CCPA, the Google Play Store requires Android developers to more responsibly communicate their apps' privacy practices to potential users by providing the proper information via the data safety, privacy policy, and permission manifest privacy transparency channels. However, it is unclear how effective those channels are in helping users make informed decisions in the app selection and installation process. In this article, we conducted a study for 190 participants to interact with our simulated privacy transparency channels of mobile apps. We quantitatively analyzed (supplemented by qualitative analysis) participants' responses to five sets of questions. We found that data safety provides the most intuitive user interfaces, privacy policy is most informative and effective, while permission manifest excels at raising participants' concerns about an app's overall privacy risks. These channels complement each other and should all be improved. △ Less

Submitted 17 November, 2025; originally announced November 2025.

Detecting the Use of Generative AI in Crowdsourced Surveys: Implications for Data Integrity

Authors: Dapeng Zhang, Marina Katoh, Weiping Pei

Abstract: The widespread adoption of generative AI (GenAI) has introduced new challenges in crowdsourced data collection, particularly in survey-based research. While GenAI offers powerful capabilities, its unintended use in crowdsourcing, such as generating automated survey responses, threatens the integrity of empirical research and complicates efforts to understand public opinion and behavior. In this st… ▽ More The widespread adoption of generative AI (GenAI) has introduced new challenges in crowdsourced data collection, particularly in survey-based research. While GenAI offers powerful capabilities, its unintended use in crowdsourcing, such as generating automated survey responses, threatens the integrity of empirical research and complicates efforts to understand public opinion and behavior. In this study, we investigate and evaluate two approaches for detecting AI-generated responses in online surveys: LLM-based detection and signature-based detection. We conducted experiments across seven survey studies, comparing responses collected before 2022 with those collected after the release of ChatGPT. Our findings reveal a significant increase in AI-generated responses in the post-2022 studies, highlighting how GenAI may silently distort crowdsourced data. This work raises broader concerns about evolving landscape of data integrity, where GenAI can compromise data quality, mislead researchers, and influence downstream findings in fields such as health, politics, and social behavior. By surfacing detection strategies and empirical evidence of GenAI's impact, we aim to contribute to ongoing conversation about safeguarding research integrity and supporting scholars navigating these methodological and ethical challenges. △ Less

Submitted 29 October, 2025; v1 submitted 28 October, 2025; originally announced October 2025.

Comments: Accepted by CSCW 2025 workshop Beyond Information: Online Participatory Culture and Information Disorder

EditInfinity: Image Editing with Binary-Quantized Generative Models

Authors: Jiahuan Wang, Yuxin Chen, Jun Yu, Guangming Lu, Wenjie Pei

Abstract: Adapting pretrained diffusion-based generative models for text-driven image editing with negligible tuning overhead has demonstrated remarkable potential. A classical adaptation paradigm, as followed by these methods, first infers the generative trajectory inversely for a given source image by image inversion, then performs image editing along the inferred trajectory guided by the target text prom… ▽ More Adapting pretrained diffusion-based generative models for text-driven image editing with negligible tuning overhead has demonstrated remarkable potential. A classical adaptation paradigm, as followed by these methods, first infers the generative trajectory inversely for a given source image by image inversion, then performs image editing along the inferred trajectory guided by the target text prompts. However, the performance of image editing is heavily limited by the approximation errors introduced during image inversion by diffusion models, which arise from the absence of exact supervision in the intermediate generative steps. To circumvent this issue, we investigate the parameter-efficient adaptation of binary-quantized generative models for image editing, and leverage their inherent characteristic that the exact intermediate quantized representations of a source image are attainable, enabling more effective supervision for precise image inversion. Specifically, we propose EditInfinity, which adapts \emph{Infinity}, a binary-quantized generative model, for image editing. We propose an efficient yet effective image inversion mechanism that integrates text prompting rectification and image style preservation, enabling precise image inversion. Furthermore, we devise a holistic smoothing strategy which allows our EditInfinity to perform image editing with high fidelity to source images and precise semantic alignment to the text prompts. Extensive experiments on the PIE-Bench benchmark across `add', `change', and `delete' editing operations, demonstrate the superior performance of our model compared to state-of-the-art diffusion-based baselines. Code available at: https://github.com/yx-chen-ust/EditInfinity. △ Less

Submitted 7 November, 2025; v1 submitted 23 October, 2025; originally announced October 2025.

Comments: 28 pages, 13 figures, accepted by The Thirty-ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025)

High-Density EEG Enables the Fastest Visual Brain-Computer Interfaces

Authors: Gege Ming, Weihua Pei, Sen Tian, Xiaogang Chen, Xiaorong Gao, Yijun Wang

Abstract: Brain-computer interface (BCI) technology establishes a direct communication pathway between the brain and external devices. Current visual BCI systems suffer from insufficient information transfer rates (ITRs) for practical use. Spatial information, a critical component of visual perception, remains underexploited in existing systems because the limited spatial resolution of recording methods hin… ▽ More Brain-computer interface (BCI) technology establishes a direct communication pathway between the brain and external devices. Current visual BCI systems suffer from insufficient information transfer rates (ITRs) for practical use. Spatial information, a critical component of visual perception, remains underexploited in existing systems because the limited spatial resolution of recording methods hinders the capture of the rich spatiotemporal dynamics of brain signals. This study proposed a frequency-phase-space fusion encoding method, integrated with 256-channel high-density electroencephalogram (EEG) recordings, to develop high-speed BCI systems. In the classical frequency-phase encoding 40-target BCI paradigm, the 256-66, 128-32, and 64-21 electrode configurations brought theoretical ITR increases of 83.66%, 79.99%, and 55.50% over the traditional 64-9 setup. In the proposed frequency-phase-space encoding 200-target BCI paradigm, these increases climbed to 195.56%, 153.08%, and 103.07%. The online BCI system achieved an average actual ITR of 472.7 bpm. This study demonstrates the essential role and immense potential of high-density EEG in decoding the spatiotemporal information of visual stimuli. △ Less

Submitted 23 July, 2025; originally announced July 2025.

MotionShot: Adaptive Motion Transfer across Arbitrary Objects for Text-to-Video Generation

Authors: Yanchen Liu, Yanan Sun, Zhening Xing, Junyao Gao, Kai Chen, Wenjie Pei

Abstract: Existing text-to-video methods struggle to transfer motion smoothly from a reference object to a target object with significant differences in appearance or structure between them. To address this challenge, we introduce MotionShot, a training-free framework capable of parsing reference-target correspondences in a fine-grained manner, thereby achieving high-fidelity motion transfer while preservin… ▽ More Existing text-to-video methods struggle to transfer motion smoothly from a reference object to a target object with significant differences in appearance or structure between them. To address this challenge, we introduce MotionShot, a training-free framework capable of parsing reference-target correspondences in a fine-grained manner, thereby achieving high-fidelity motion transfer while preserving coherence in appearance. To be specific, MotionShot first performs semantic feature matching to ensure high-level alignments between the reference and target objects. It then further establishes low-level morphological alignments through reference-to-target shape retargeting. By encoding motion with temporal attention, our MotionShot can coherently transfer motion across objects, even in the presence of significant appearance and structure disparities, demonstrated by extensive experiments. The project page is available at: https://motionshot.github.io/. △ Less

Submitted 22 July, 2025; originally announced July 2025.

Enhancing Spatial Reasoning in Multimodal Large Language Models through Reasoning-based Segmentation

Authors: Zhenhua Ning, Zhuotao Tian, Shaoshuai Shi, Guangming Lu, Daojing He, Wenjie Pei, Li Jiang

Abstract: Recent advances in point cloud perception have demonstrated remarkable progress in scene understanding through vision-language alignment leveraging large language models (LLMs). However, existing methods may still encounter challenges in handling complex instructions that require accurate spatial reasoning, even if the 3D point cloud data provides detailed spatial cues such as size and position fo… ▽ More Recent advances in point cloud perception have demonstrated remarkable progress in scene understanding through vision-language alignment leveraging large language models (LLMs). However, existing methods may still encounter challenges in handling complex instructions that require accurate spatial reasoning, even if the 3D point cloud data provides detailed spatial cues such as size and position for identifying the targets. To tackle this issue, we propose Relevant Reasoning Segmentation (R$^2$S), a reasoning-based segmentation framework. The framework emulates human cognitive processes by decomposing spatial reasoning into two sequential stages: first identifying relevant elements, then processing instructions guided by their associated visual priors. Furthermore, acknowledging the inadequacy of existing datasets in complex reasoning tasks, we introduce 3D ReasonSeg, a reasoning-based segmentation dataset comprising 25,185 training samples and 3,966 validation samples with precise annotations. Both quantitative and qualitative experiments demonstrate that the R$^2$S and 3D ReasonSeg effectively endow 3D point cloud perception with stronger spatial reasoning capabilities, and we hope that they can serve as a new baseline and benchmark for future work. △ Less

Submitted 29 June, 2025; originally announced June 2025.

Lightweight Joint Audio-Visual Deepfake Detection via Single-Stream Multi-Modal Learning Framework

Authors: Kuiyuan Zhang, Wenjie Pei, Rushi Lan, Yifang Guo, Zhongyun Hua

Abstract: Deepfakes are AI-synthesized multimedia data that may be abused for spreading misinformation. Deepfake generation involves both visual and audio manipulation. To detect audio-visual deepfakes, previous studies commonly employ two relatively independent sub-models to learn audio and visual features, respectively, and fuse them subsequently for deepfake detection. However, this may underutilize the… ▽ More Deepfakes are AI-synthesized multimedia data that may be abused for spreading misinformation. Deepfake generation involves both visual and audio manipulation. To detect audio-visual deepfakes, previous studies commonly employ two relatively independent sub-models to learn audio and visual features, respectively, and fuse them subsequently for deepfake detection. However, this may underutilize the inherent correlations between audio and visual features. Moreover, utilizing two isolated feature learning sub-models can result in redundant neural layers, making the overall model inefficient and impractical for resource-constrained environments. In this work, we design a lightweight network for audio-visual deepfake detection via a single-stream multi-modal learning framework. Specifically, we introduce a collaborative audio-visual learning block to efficiently integrate multi-modal information while learning the visual and audio features. By iteratively employing this block, our single-stream network achieves a continuous fusion of multi-modal features across its layers. Thus, our network efficiently captures visual and audio features without the need for excessive block stacking, resulting in a lightweight network design. Furthermore, we propose a multi-modal classification module that can boost the dependence of the visual and audio classifiers on modality content. It also enhances the whole resistance of the video classifier against the mismatches between audio and visual modalities. We conduct experiments on the DF-TIMIT, FakeAVCeleb, and DFDC benchmark datasets. Compared to state-of-the-art audio-visual joint detection methods, our method is significantly lightweight with only 0.48M parameters, yet it achieves superiority in both uni-modal and multi-modal deepfakes, as well as in unseen types of deepfakes. △ Less

Submitted 8 June, 2025; originally announced June 2025.

Learning Compatible Multi-Prize Subnetworks for Asymmetric Retrieval

Authors: Yushuai Sun, Zikun Zhou, Dongmei Jiang, Yaowei Wang, Jun Yu, Guangming Lu, Wenjie Pei

Abstract: Asymmetric retrieval is a typical scenario in real-world retrieval systems, where compatible models of varying capacities are deployed on platforms with different resource configurations. Existing methods generally train pre-defined networks or subnetworks with capacities specifically designed for pre-determined platforms, using compatible learning. Nevertheless, these methods suffer from limited… ▽ More Asymmetric retrieval is a typical scenario in real-world retrieval systems, where compatible models of varying capacities are deployed on platforms with different resource configurations. Existing methods generally train pre-defined networks or subnetworks with capacities specifically designed for pre-determined platforms, using compatible learning. Nevertheless, these methods suffer from limited flexibility for multi-platform deployment. For example, when introducing a new platform into the retrieval systems, developers have to train an additional model at an appropriate capacity that is compatible with existing models via backward-compatible learning. In this paper, we propose a Prunable Network with self-compatibility, which allows developers to generate compatible subnetworks at any desired capacity through post-training pruning. Thus it allows the creation of a sparse subnetwork matching the resources of the new platform without additional training. Specifically, we optimize both the architecture and weight of subnetworks at different capacities within a dense network in compatible learning. We also design a conflict-aware gradient integration scheme to handle the gradient conflicts between the dense network and subnetworks during compatible learning. Extensive experiments on diverse benchmarks and visual backbones demonstrate the effectiveness of our method. Our code and model are available at https://github.com/Bunny-Black/PrunNet. △ Less

Submitted 16 April, 2025; originally announced April 2025.

Comments: Accepted to CVPR 2025

Feather-SQL: A Lightweight NL2SQL Framework with Dual-Model Collaboration Paradigm for Small Language Models

Authors: Wenqi Pei, Hailing Xu, Hengyuan Zhao, Shizheng Hou, Han Chen, Zining Zhang, Pingyi Luo, Bingsheng He

Abstract: Natural Language to SQL (NL2SQL) has seen significant advancements with large language models (LLMs). However, these models often depend on closed-source systems and high computational resources, posing challenges in data privacy and deployment. In contrast, small language models (SLMs) struggle with NL2SQL tasks, exhibiting poor performance and incompatibility with existing frameworks. To address… ▽ More Natural Language to SQL (NL2SQL) has seen significant advancements with large language models (LLMs). However, these models often depend on closed-source systems and high computational resources, posing challenges in data privacy and deployment. In contrast, small language models (SLMs) struggle with NL2SQL tasks, exhibiting poor performance and incompatibility with existing frameworks. To address these issues, we introduce Feather-SQL, a new lightweight framework tailored for SLMs. Feather-SQL improves SQL executability and accuracy through 1) schema pruning and linking, 2) multi-path and multi-candidate generation. Additionally, we introduce the 1+1 Model Collaboration Paradigm, which pairs a strong general-purpose chat model with a fine-tuned SQL specialist, combining strong analytical reasoning with high-precision SQL generation. Experimental results on BIRD demonstrate that Feather-SQL improves NL2SQL performance on SLMs, with around 10% boost for models without fine-tuning. The proposed paradigm raises the accuracy ceiling of SLMs to 54.76%, highlighting its effectiveness. △ Less

Submitted 18 August, 2025; v1 submitted 22 March, 2025; originally announced March 2025.

Comments: DL4C @ ICLR 2025

Prototype Perturbation for Relaxing Alignment Constraints in Backward-Compatible Learning

Authors: Zikun Zhou, Yushuai Sun, Wenjie Pei, Xin Li, Yaowei Wang

Abstract: The traditional paradigm to update retrieval models requires re-computing the embeddings of the gallery data, a time-consuming and computationally intensive process known as backfilling. To circumvent backfilling, Backward-Compatible Learning (BCL) has been widely explored, which aims to train a new model compatible with the old one. Many previous works focus on effectively aligning the embeddings… ▽ More The traditional paradigm to update retrieval models requires re-computing the embeddings of the gallery data, a time-consuming and computationally intensive process known as backfilling. To circumvent backfilling, Backward-Compatible Learning (BCL) has been widely explored, which aims to train a new model compatible with the old one. Many previous works focus on effectively aligning the embeddings of the new model with those of the old one to enhance the backward-compatibility. Nevertheless, such strong alignment constraints would compromise the discriminative ability of the new model, particularly when different classes are closely clustered and hard to distinguish in the old feature space. To address this issue, we propose to relax the constraints by introducing perturbations to the old feature prototypes. This allows us to align the new feature space with a pseudo-old feature space defined by these perturbed prototypes, thereby preserving the discriminative ability of the new model in backward-compatible learning. We have developed two approaches for calculating the perturbations: Neighbor-Driven Prototype Perturbation (NDPP) and Optimization-Driven Prototype Perturbation (ODPP). Particularly, they take into account the feature distributions of not only the old but also the new models to obtain proper perturbations along with new model updating. Extensive experiments on the landmark and commodity datasets demonstrate that our approaches perform favorably against state-of-the-art BCL algorithms. △ Less

Submitted 7 March, 2026; v1 submitted 18 March, 2025; originally announced March 2025.

Comments: Accept to IEEE TMM

Recognition-Synergistic Scene Text Editing

Authors: Zhengyao Fang, Pengyuan Lyu, Jingjing Wu, Chengquan Zhang, Jun Yu, Guangming Lu, Wenjie Pei

Abstract: Scene text editing aims to modify text content within scene images while maintaining style consistency. Traditional methods achieve this by explicitly disentangling style and content from the source image and then fusing the style with the target content, while ensuring content consistency using a pre-trained recognition model. Despite notable progress, these methods suffer from complex pipelines,… ▽ More Scene text editing aims to modify text content within scene images while maintaining style consistency. Traditional methods achieve this by explicitly disentangling style and content from the source image and then fusing the style with the target content, while ensuring content consistency using a pre-trained recognition model. Despite notable progress, these methods suffer from complex pipelines, leading to suboptimal performance in complex scenarios. In this work, we introduce Recognition-Synergistic Scene Text Editing (RS-STE), a novel approach that fully exploits the intrinsic synergy of text recognition for editing. Our model seamlessly integrates text recognition with text editing within a unified framework, and leverages the recognition model's ability to implicitly disentangle style and content while ensuring content consistency. Specifically, our approach employs a multi-modal parallel decoder based on transformer architecture, which predicts both text content and stylized images in parallel. Additionally, our cyclic self-supervised fine-tuning strategy enables effective training on unpaired real-world data without ground truth, enhancing style and content consistency through a twice-cyclic generation process. Built on a relatively simple architecture, RS-STE achieves state-of-the-art performance on both synthetic and real-world benchmarks, and further demonstrates the effectiveness of leveraging the generated hard cases to boost the performance of downstream recognition tasks. Code is available at https://github.com/ZhengyaoFang/RS-STE. △ Less

Submitted 10 March, 2026; v1 submitted 11 March, 2025; originally announced March 2025.

Comments: accepted by CVPR2025

InterFeedback: Unveiling Interactive Intelligence of Large Multimodal Models via Human Feedback

Authors: Henry Hengyuan Zhao, Wenqi Pei, Yifei Tao, Haiyang Mei, Mike Zheng Shou

Abstract: Existing benchmarks do not test Large Multimodal Models (LMMs) on their interactive intelligence with human users, which is vital for developing general-purpose AI assistants. We design InterFeedback, an interactive framework, which can be applied to any LMM and dataset to assess this ability autonomously. On top of this, we introduce InterFeedback-Bench which evaluates interactive intelligence us… ▽ More Existing benchmarks do not test Large Multimodal Models (LMMs) on their interactive intelligence with human users, which is vital for developing general-purpose AI assistants. We design InterFeedback, an interactive framework, which can be applied to any LMM and dataset to assess this ability autonomously. On top of this, we introduce InterFeedback-Bench which evaluates interactive intelligence using two representative datasets, MMMU-Pro and MathVerse, to test 10 different open-source LMMs. Additionally, we present InterFeedback-Human, a newly collected dataset of 120 cases designed for manually testing interactive performance in leading models such as OpenAI-o1 and Claude-Sonnet-4. Our evaluation results indicate that even the state-of-the-art LMM, OpenAI-o1, struggles to refine its responses based on human feedback, achieving an average score of less than 50%. Our findings point to the need for methods that can enhance LMMs' capabilities to interpret and benefit from feedback. △ Less

Submitted 7 November, 2025; v1 submitted 20 February, 2025; originally announced February 2025.

Comments: Accepted by EMNLP 2025 Findings

Federated Unlearning Model Recovery in Data with Skewed Label Distributions

Authors: Xinrui Yu, Wenbin Pei, Bing Xue, Qiang Zhang

Abstract: In federated learning, federated unlearning is a technique that provides clients with a rollback mechanism that allows them to withdraw their data contribution without training from scratch. However, existing research has not considered scenarios with skewed label distributions. Unfortunately, the unlearning of a client with skewed data usually results in biased models and makes it difficult to de… ▽ More In federated learning, federated unlearning is a technique that provides clients with a rollback mechanism that allows them to withdraw their data contribution without training from scratch. However, existing research has not considered scenarios with skewed label distributions. Unfortunately, the unlearning of a client with skewed data usually results in biased models and makes it difficult to deliver high-quality service, complicating the recovery process. This paper proposes a recovery method of federated unlearning with skewed label distributions. Specifically, we first adopt a strategy that incorporates oversampling with deep learning to supplement the skewed class data for clients to perform recovery training, therefore enhancing the completeness of their local datasets. Afterward, a density-based denoising method is applied to remove noise from the generated data, further improving the quality of the remaining clients' datasets. Finally, all the remaining clients leverage the enhanced local datasets and engage in iterative training to effectively restore the performance of the unlearning model. Extensive evaluations on commonly used federated learning datasets with varying degrees of skewness show that our method outperforms baseline methods in restoring the performance of the unlearning model, particularly regarding accuracy on the skewed class. △ Less

Submitted 20 December, 2024; v1 submitted 17 December, 2024; originally announced December 2024.

EvoSampling: A Granular Ball-based Evolutionary Hybrid Sampling with Knowledge Transfer for Imbalanced Learning

Authors: Wenbin Pei, Ruohao Dai, Bing Xue, Mengjie Zhang, Qiang Zhang, Yiu-Ming Cheung, Shuyin Xia

Abstract: Class imbalance would lead to biased classifiers that favor the majority class and disadvantage the minority class. Unfortunately, from a practical perspective, the minority class is of importance in many real-life applications. Hybrid sampling methods address this by oversampling the minority class to increase the number of its instances, followed by undersampling to remove low-quality instances.… ▽ More Class imbalance would lead to biased classifiers that favor the majority class and disadvantage the minority class. Unfortunately, from a practical perspective, the minority class is of importance in many real-life applications. Hybrid sampling methods address this by oversampling the minority class to increase the number of its instances, followed by undersampling to remove low-quality instances. However, most existing sampling methods face difficulties in generating diverse high-quality instances and often fail to remove noise or low-quality instances on a larger scale effectively. This paper therefore proposes an evolutionary multi-granularity hybrid sampling method, called EvoSampling. During the oversampling process, genetic programming (GP) is used with multi-task learning to effectively and efficiently generate diverse high-quality instances. During the undersampling process, we develop a granular ball-based undersampling method that removes noise in a multi-granular fashion, thereby enhancing data quality. Experiments on 20 imbalanced datasets demonstrate that EvoSampling effectively enhances the performance of various classification algorithms by providing better datasets than existing sampling methods. Besides, ablation studies further indicate that allowing knowledge transfer accelerates the GP's evolutionary learning process. △ Less

Submitted 12 December, 2024; originally announced December 2024.

UmambaTSF: A U-shaped Multi-Scale Long-Term Time Series Forecasting Method Using Mamba

Authors: Li Wu, Wenbin Pei, Jiulong Jiao, Qiang Zhang

Abstract: Multivariate Time series forecasting is crucial in domains such as transportation, meteorology, and finance, especially for predicting extreme weather events. State-of-the-art methods predominantly rely on Transformer architectures, which utilize attention mechanisms to capture temporal dependencies. However, these methods are hindered by quadratic time complexity, limiting the model's scalability… ▽ More Multivariate Time series forecasting is crucial in domains such as transportation, meteorology, and finance, especially for predicting extreme weather events. State-of-the-art methods predominantly rely on Transformer architectures, which utilize attention mechanisms to capture temporal dependencies. However, these methods are hindered by quadratic time complexity, limiting the model's scalability with respect to input sequence length. This significantly restricts their practicality in the real world. Mamba, based on state space models (SSM), provides a solution with linear time complexity, increasing the potential for efficient forecasting of sequential data. In this study, we propose UmambaTSF, a novel long-term time series forecasting framework that integrates multi-scale feature extraction capabilities of U-shaped encoder-decoder multilayer perceptrons (MLP) with Mamba's long sequence representation. To improve performance and efficiency, the Mamba blocks introduced in the framework adopt a refined residual structure and adaptable design, enabling the capture of unique temporal signals and flexible channel processing. In the experiments, UmambaTSF achieves state-of-the-art performance and excellent generality on widely used benchmark datasets while maintaining linear time complexity and low memory consumption. △ Less

Submitted 15 October, 2024; originally announced October 2024.

DivDiff: A Conditional Diffusion Model for Diverse Human Motion Prediction

Authors: Hua Yu, Yaqing Hou, Wenbin Pei, Qiang Zhang

Abstract: Diverse human motion prediction (HMP) aims to predict multiple plausible future motions given an observed human motion sequence. It is a challenging task due to the diversity of potential human motions while ensuring an accurate description of future human motions. Current solutions are either low-diversity or limited in expressiveness. Recent denoising diffusion models (DDPM) hold potential gener… ▽ More Diverse human motion prediction (HMP) aims to predict multiple plausible future motions given an observed human motion sequence. It is a challenging task due to the diversity of potential human motions while ensuring an accurate description of future human motions. Current solutions are either low-diversity or limited in expressiveness. Recent denoising diffusion models (DDPM) hold potential generative capabilities in generative tasks. However, introducing DDPM directly into diverse HMP incurs some issues. Although DDPM can increase the diversity of the potential patterns of human motions, the predicted human motions become implausible over time because of the significant noise disturbances in the forward process of DDPM. This phenomenon leads to the predicted human motions being hard to control, seriously impacting the quality of predicted motions and restricting their practical applicability in real-world scenarios. To alleviate this, we propose a novel conditional diffusion-based generative model, called DivDiff, to predict more diverse and realistic human motions. Specifically, the DivDiff employs DDPM as our backbone and incorporates Discrete Cosine Transform (DCT) and transformer mechanisms to encode the observed human motion sequence as a condition to instruct the reverse process of DDPM. More importantly, we design a diversified reinforcement sampling function (DRSF) to enforce human skeletal constraints on the predicted human motions. DRSF utilizes the acquired information from human skeletal as prior knowledge, thereby reducing significant disturbances introduced during the forward process. Extensive results received in the experiments on two widely-used datasets (Human3.6M and HumanEva-I) demonstrate that our model obtains competitive performance on both diversity and accuracy. △ Less

Submitted 16 August, 2024; originally announced September 2024.

SynopGround: A Large-Scale Dataset for Multi-Paragraph Video Grounding from TV Dramas and Synopses

Authors: Chaolei Tan, Zihang Lin, Junfu Pu, Zhongang Qi, Wei-Yi Pei, Zhi Qu, Yexin Wang, Ying Shan, Wei-Shi Zheng, Jian-Fang Hu

Abstract: Video grounding is a fundamental problem in multimodal content understanding, aiming to localize specific natural language queries in an untrimmed video. However, current video grounding datasets merely focus on simple events and are either limited to shorter videos or brief sentences, which hinders the model from evolving toward stronger multimodal understanding capabilities. To address these lim… ▽ More Video grounding is a fundamental problem in multimodal content understanding, aiming to localize specific natural language queries in an untrimmed video. However, current video grounding datasets merely focus on simple events and are either limited to shorter videos or brief sentences, which hinders the model from evolving toward stronger multimodal understanding capabilities. To address these limitations, we present a large-scale video grounding dataset named SynopGround, in which more than 2800 hours of videos are sourced from popular TV dramas and are paired with accurately localized human-written synopses. Each paragraph in the synopsis serves as a language query and is manually annotated with precise temporal boundaries in the long video. These paragraph queries are tightly correlated to each other and contain a wealth of abstract expressions summarizing video storylines and specific descriptions portraying event details, which enables the model to learn multimodal perception on more intricate concepts over longer context dependencies. Based on the dataset, we further introduce a more complex setting of video grounding dubbed Multi-Paragraph Video Grounding (MPVG), which takes as input multiple paragraphs and a long video for grounding each paragraph query to its temporal interval. In addition, we propose a novel Local-Global Multimodal Reasoner (LGMR) to explicitly model the local-global structures of long-term multimodal inputs for MPVG. Our method provides an effective baseline solution to the multi-paragraph video grounding problem. Extensive experiments verify the proposed model's effectiveness as well as its superiority in long-term multi-paragraph video grounding over prior state-of-the-arts. Dataset and code are publicly available. Project page: https://synopground.github.io/. △ Less

Submitted 18 August, 2024; v1 submitted 3 August, 2024; originally announced August 2024.

Comments: Accepted to ACM MM 2024. Project page: https://synopground.github.io/

UniVoxel: Fast Inverse Rendering by Unified Voxelization of Scene Representation

Authors: Shuang Wu, Songlin Tang, Guangming Lu, Jianzhuang Liu, Wenjie Pei

Abstract: Typical inverse rendering methods focus on learning implicit neural scene representations by modeling the geometry, materials and illumination separately, which entails significant computations for optimization. In this work we design a Unified Voxelization framework for explicit learning of scene representations, dubbed UniVoxel, which allows for efficient modeling of the geometry, materials and… ▽ More Typical inverse rendering methods focus on learning implicit neural scene representations by modeling the geometry, materials and illumination separately, which entails significant computations for optimization. In this work we design a Unified Voxelization framework for explicit learning of scene representations, dubbed UniVoxel, which allows for efficient modeling of the geometry, materials and illumination jointly, thereby accelerating the inverse rendering significantly. To be specific, we propose to encode a scene into a latent volumetric representation, based on which the geometry, materials and illumination can be readily learned via lightweight neural networks in a unified manner. Particularly, an essential design of UniVoxel is that we leverage local Spherical Gaussians to represent the incident light radiance, which enables the seamless integration of modeling illumination into the unified voxelization framework. Such novel design enables our UniVoxel to model the joint effects of direct lighting, indirect lighting and light visibility efficiently without expensive multi-bounce ray tracing. Extensive experiments on multiple benchmarks covering diverse scenes demonstrate that UniVoxel boosts the optimization efficiency significantly compared to other methods, reducing the per-scene training time from hours to 18 minutes, while achieving favorable reconstruction quality. Code is available at https://github.com/freemantom/UniVoxel. △ Less

Submitted 28 July, 2024; originally announced July 2024.

Comments: ECCV2024

WeCromCL: Weakly Supervised Cross-Modality Contrastive Learning for Transcription-only Supervised Text Spotting

Authors: Jingjing Wu, Zhengyao Fang, Pengyuan Lyu, Chengquan Zhang, Fanglin Chen, Guangming Lu, Wenjie Pei

Abstract: Transcription-only Supervised Text Spotting aims to learn text spotters relying only on transcriptions but no text boundaries for supervision, thus eliminating expensive boundary annotation. The crux of this task lies in locating each transcription in scene text images without location annotations. In this work, we formulate this challenging problem as a Weakly Supervised Cross-modality Contrastiv… ▽ More Transcription-only Supervised Text Spotting aims to learn text spotters relying only on transcriptions but no text boundaries for supervision, thus eliminating expensive boundary annotation. The crux of this task lies in locating each transcription in scene text images without location annotations. In this work, we formulate this challenging problem as a Weakly Supervised Cross-modality Contrastive Learning problem, and design a simple yet effective model dubbed WeCromCL that is able to detect each transcription in a scene image in a weakly supervised manner. Unlike typical methods for cross-modality contrastive learning that focus on modeling the holistic semantic correlation between an entire image and a text description, our WeCromCL conducts atomistic contrastive learning to model the character-wise appearance consistency between a text transcription and its correlated region in a scene image to detect an anchor point for the transcription in a weakly supervised manner. The detected anchor points by WeCromCL are further used as pseudo location labels to guide the learning of text spotting. Extensive experiments on four challenging benchmarks demonstrate the superior performance of our model over other methods. Code will be released. △ Less

Submitted 13 January, 2025; v1 submitted 28 July, 2024; originally announced July 2024.

Comments: Accepted by ECCV 2024

AnyControl: Create Your Artwork with Versatile Control on Text-to-Image Generation

Authors: Yanan Sun, Yanchen Liu, Yinhao Tang, Wenjie Pei, Kai Chen

Abstract: The field of text-to-image (T2I) generation has made significant progress in recent years, largely driven by advancements in diffusion models. Linguistic control enables effective content creation, but struggles with fine-grained control over image generation. This challenge has been explored, to a great extent, by incorporating additional user-supplied spatial conditions, such as depth maps and e… ▽ More The field of text-to-image (T2I) generation has made significant progress in recent years, largely driven by advancements in diffusion models. Linguistic control enables effective content creation, but struggles with fine-grained control over image generation. This challenge has been explored, to a great extent, by incorporating additional user-supplied spatial conditions, such as depth maps and edge maps, into pre-trained T2I models through extra encoding. However, multi-control image synthesis still faces several challenges. Specifically, current approaches are limited in handling free combinations of diverse input control signals, overlook the complex relationships among multiple spatial conditions, and often fail to maintain semantic alignment with provided textual prompts. This can lead to suboptimal user experiences. To address these challenges, we propose AnyControl, a multi-control image synthesis framework that supports arbitrary combinations of diverse control signals. AnyControl develops a novel Multi-Control Encoder that extracts a unified multi-modal embedding to guide the generation process. This approach enables a holistic understanding of user inputs, and produces high-quality, faithful results under versatile control signals, as demonstrated by extensive quantitative and qualitative evaluations. Our project page is available in https://any-control.github.io. △ Less

Submitted 18 July, 2024; v1 submitted 27 June, 2024; originally announced June 2024.

Comments: Accepted by ECCV 2024, code and dataset available in https://github.com/open-mmlab/AnyControl

Influence Maximization in Hypergraphs Using A Genetic Algorithm with New Initialization and Evaluation Methods

Authors: Xilong Qu, Wenbin Pei, Yingchao Yang, Xirong Xu, Renquan Zhang, Qiang Zhang

Abstract: Influence maximization (IM) is a crucial optimization task related to analyzing complex networks in the real world, such as social networks, disease propagation networks, and marketing networks. Publications to date about the IM problem focus mainly on graphs, which fail to capture high-order interaction relationships from the real world. Therefore, the use of hypergraphs for addressing the IM pro… ▽ More Influence maximization (IM) is a crucial optimization task related to analyzing complex networks in the real world, such as social networks, disease propagation networks, and marketing networks. Publications to date about the IM problem focus mainly on graphs, which fail to capture high-order interaction relationships from the real world. Therefore, the use of hypergraphs for addressing the IM problem has been receiving increasing attention. However, identifying the most influential nodes in hypergraphs remains challenging, mainly because nodes and hyperedges are often strongly coupled and correlated. In this paper, to effectively identify the most influential nodes, we first propose a novel hypergraph-independent cascade model that integrates the influences of both node and hyperedge failures. Afterward, we introduce genetic algorithms (GA) to identify the most influential nodes that leverage hypergraph collective influences. In the GA-based method, the hypergraph collective influence is effectively used to initialize the population, thereby enhancing the quality of initial candidate solutions. The designed fitness function considers the joint influences of both nodes and hyperedges. This ensures the optimal set of nodes with the best influence on both nodes and hyperedges to be evaluated accurately. Moreover, a new mutation operator is designed by introducing factors, i.e., the collective influence and overlapping effects of nodes in hypergraphs, to breed high-quality offspring. In the experiments, several simulations on both synthetic and real hypergraphs have been conducted, and the results demonstrate that the proposed method outperforms the compared methods. △ Less

Submitted 15 May, 2024; originally announced May 2024.

Domain-Rectifying Adapter for Cross-Domain Few-Shot Segmentation

Authors: Jiapeng Su, Qi Fan, Guangming Lu, Fanglin Chen, Wenjie Pei

Abstract: Few-shot semantic segmentation (FSS) has achieved great success on segmenting objects of novel classes, supported by only a few annotated samples. However, existing FSS methods often underperform in the presence of domain shifts, especially when encountering new domain styles that are unseen during training. It is suboptimal to directly adapt or generalize the entire model to new domains in the fe… ▽ More Few-shot semantic segmentation (FSS) has achieved great success on segmenting objects of novel classes, supported by only a few annotated samples. However, existing FSS methods often underperform in the presence of domain shifts, especially when encountering new domain styles that are unseen during training. It is suboptimal to directly adapt or generalize the entire model to new domains in the few-shot scenario. Instead, our key idea is to adapt a small adapter for rectifying diverse target domain styles to the source domain. Consequently, the rectified target domain features can fittingly benefit from the well-optimized source domain segmentation model, which is intently trained on sufficient source domain data. Training domain-rectifying adapter requires sufficiently diverse target domains. We thus propose a novel local-global style perturbation method to simulate diverse potential target domains by perturbating the feature channel statistics of the individual images and collective statistics of the entire source domain, respectively. Additionally, we propose a cyclic domain alignment module to facilitate the adapter effectively rectifying domains using a reverse domain rectification supervision. The adapter is trained to rectify the image features from diverse synthesized target domains to align with the source domain. During testing on target domains, we start by rectifying the image features and then conduct few-shot segmentation on the domain-rectified features. Extensive experiments demonstrate the effectiveness of our method, achieving promising results on cross-domain few-shot semantic segmentation tasks. Our code is available at https://github.com/Matt-Su/DR-Adapter. △ Less

Submitted 16 April, 2024; originally announced April 2024.

Comments: Accepted by CVPR 2024

Improving Critical Node Detection Using Neural Network-based Initialization in a Genetic Algorithm

Authors: Chanjuan Liu, Shike Ge, Zhihan Chen, Wenbin Pei, Enqiang Zhu, Yi Mei, Hisao Ishibuchi

Abstract: The Critical Node Problem (CNP) is concerned with identifying the critical nodes in a complex network. These nodes play a significant role in maintaining the connectivity of the network, and removing them can negatively impact network performance. CNP has been studied extensively due to its numerous real-world applications. Among the different versions of CNP, CNP-1a has gained the most popularity… ▽ More The Critical Node Problem (CNP) is concerned with identifying the critical nodes in a complex network. These nodes play a significant role in maintaining the connectivity of the network, and removing them can negatively impact network performance. CNP has been studied extensively due to its numerous real-world applications. Among the different versions of CNP, CNP-1a has gained the most popularity. The primary objective of CNP-1a is to minimize the pair-wise connectivity in the remaining network after deleting a limited number of nodes from a network. Due to the NP-hard nature of CNP-1a, many heuristic/metaheuristic algorithms have been proposed to solve this problem. However, most existing algorithms start with a random initialization, leading to a high cost of obtaining an optimal solution. To improve the efficiency of solving CNP-1a, a knowledge-guided genetic algorithm named K2GA has been proposed. Unlike the standard genetic algorithm framework, K2GA has two main components: a pretrained neural network to obtain prior knowledge on possible critical nodes, and a hybrid genetic algorithm with local search for finding an optimal set of critical nodes based on the knowledge given by the trained neural network. The local search process utilizes a cut node-based greedy strategy. The effectiveness of the proposed knowledgeguided genetic algorithm is verified by experiments on 26 realworld instances of complex networks. Experimental results show that K2GA outperforms the state-of-the-art algorithms regarding the best, median, and average objective values, and improves the best upper bounds on the best objective values for eight realworld instances. △ Less

Submitted 1 February, 2024; originally announced February 2024.

Comments: 14 pages, 13 figures

Saliency-Aware Regularized Graph Neural Network

Authors: Wenjie Pei, Weina Xu, Zongze Wu, Weichao Li, Jinfan Wang, Guangming Lu, Xiangrong Wang

Abstract: The crux of graph classification lies in the effective representation learning for the entire graph. Typical graph neural networks focus on modeling the local dependencies when aggregating features of neighboring nodes, and obtain the representation for the entire graph by aggregating node features. Such methods have two potential limitations: 1) the global node saliency w.r.t. graph classificatio… ▽ More The crux of graph classification lies in the effective representation learning for the entire graph. Typical graph neural networks focus on modeling the local dependencies when aggregating features of neighboring nodes, and obtain the representation for the entire graph by aggregating node features. Such methods have two potential limitations: 1) the global node saliency w.r.t. graph classification is not explicitly modeled, which is crucial since different nodes may have different semantic relevance to graph classification; 2) the graph representation directly aggregated from node features may have limited effectiveness to reflect graph-level information. In this work, we propose the Saliency-Aware Regularized Graph Neural Network (SAR-GNN) for graph classification, which consists of two core modules: 1) a traditional graph neural network serving as the backbone for learning node features and 2) the Graph Neural Memory designed to distill a compact graph representation from node features of the backbone. We first estimate the global node saliency by measuring the semantic similarity between the compact graph representation and node features. Then the learned saliency distribution is leveraged to regularize the neighborhood aggregation of the backbone, which facilitates the message passing of features for salient nodes and suppresses the less relevant nodes. Thus, our model can learn more effective graph representation. We demonstrate the merits of SAR-GNN by extensive experiments on seven datasets across various types of graph data. Code will be released. △ Less

Submitted 1 January, 2024; originally announced January 2024.

Comments: Accepted by Artificial Intelligence Journal with minor revision

Robust 3D Tracking with Quality-Aware Shape Completion

Authors: Jingwen Zhang, Zikun Zhou, Guangming Lu, Jiandong Tian, Wenjie Pei

Abstract: 3D single object tracking remains a challenging problem due to the sparsity and incompleteness of the point clouds. Existing algorithms attempt to address the challenges in two strategies. The first strategy is to learn dense geometric features based on the captured sparse point cloud. Nevertheless, it is quite a formidable task since the learned dense geometric features are with high uncertainty… ▽ More 3D single object tracking remains a challenging problem due to the sparsity and incompleteness of the point clouds. Existing algorithms attempt to address the challenges in two strategies. The first strategy is to learn dense geometric features based on the captured sparse point cloud. Nevertheless, it is quite a formidable task since the learned dense geometric features are with high uncertainty for depicting the shape of the target object. The other strategy is to aggregate the sparse geometric features of multiple templates to enrich the shape information, which is a routine solution in 2D tracking. However, aggregating the coarse shape representations can hardly yield a precise shape representation. Different from 2D pixels, 3D points of different frames can be directly fused by coordinate transform, i.e., shape completion. Considering that, we propose to construct a synthetic target representation composed of dense and complete point clouds depicting the target shape precisely by shape completion for robust 3D tracking. Specifically, we design a voxelized 3D tracking framework with shape completion, in which we propose a quality-aware shape completion mechanism to alleviate the adverse effect of noisy historical predictions. It enables us to effectively construct and leverage the synthetic target representation. Besides, we also develop a voxelized relation modeling module and box refinement module to improve tracking performance. Favorable performance against state-of-the-art algorithms on three benchmarks demonstrates the effectiveness and generalization ability of our method. △ Less

Submitted 16 December, 2023; originally announced December 2023.

Comments: A detailed version of the paper accepted by AAAI 2024

SA$^2$VP: Spatially Aligned-and-Adapted Visual Prompt

Authors: Wenjie Pei, Tongqi Xia, Fanglin Chen, Jinsong Li, Jiandong Tian, Guangming Lu

Abstract: As a prominent parameter-efficient fine-tuning technique in NLP, prompt tuning is being explored its potential in computer vision. Typical methods for visual prompt tuning follow the sequential modeling paradigm stemming from NLP, which represents an input image as a flattened sequence of token embeddings and then learns a set of unordered parameterized tokens prefixed to the sequence representati… ▽ More As a prominent parameter-efficient fine-tuning technique in NLP, prompt tuning is being explored its potential in computer vision. Typical methods for visual prompt tuning follow the sequential modeling paradigm stemming from NLP, which represents an input image as a flattened sequence of token embeddings and then learns a set of unordered parameterized tokens prefixed to the sequence representation as the visual prompts for task adaptation of large vision models. While such sequential modeling paradigm of visual prompt has shown great promise, there are two potential limitations. First, the learned visual prompts cannot model the underlying spatial relations in the input image, which is crucial for image encoding. Second, since all prompt tokens play the same role of prompting for all image tokens without distinction, it lacks the fine-grained prompting capability, i.e., individual prompting for different image tokens. In this work, we propose the \mymodel model (\emph{SA$^2$VP}), which learns a two-dimensional prompt token map with equal (or scaled) size to the image token map, thereby being able to spatially align with the image map. Each prompt token is designated to prompt knowledge only for the spatially corresponding image tokens. As a result, our model can conduct individual prompting for different image tokens in a fine-grained manner. Moreover, benefiting from the capability of preserving the spatial structure by the learned prompt token map, our \emph{SA$^2$VP} is able to model the spatial relations in the input image, leading to more effective prompting. Extensive experiments on three challenging benchmarks for image classification demonstrate the superiority of our model over other state-of-the-art methods for visual prompt tuning. Code is available at \emph{https://github.com/tommy-xq/SA2VP}. △ Less

Submitted 16 December, 2023; originally announced December 2023.

Comments: Accepted by AAAI 2024

D$^2$ST-Adapter: Disentangled-and-Deformable Spatio-Temporal Adapter for Few-shot Action Recognition

Authors: Wenjie Pei, Qizhong Tan, Guangming Lu, Jiandong Tian, Jun Yu

Abstract: Adapting pre-trained image models to video modality has proven to be an effective strategy for robust few-shot action recognition. In this work, we explore the potential of adapter tuning in image-to-video model adaptation and propose a novel video adapter tuning framework, called Disentangled-and-Deformable Spatio-Temporal Adapter (D$^2$ST-Adapter). It features a lightweight design, low adaptatio… ▽ More Adapting pre-trained image models to video modality has proven to be an effective strategy for robust few-shot action recognition. In this work, we explore the potential of adapter tuning in image-to-video model adaptation and propose a novel video adapter tuning framework, called Disentangled-and-Deformable Spatio-Temporal Adapter (D$^2$ST-Adapter). It features a lightweight design, low adaptation overhead and powerful spatio-temporal feature adaptation capabilities. D$^2$ST-Adapter is structured with an internal dual-pathway architecture that enables built-in disentangled encoding of spatial and temporal features within the adapter, seamlessly integrating into the single-stream feature learning framework of pre-trained image models. In particular, we develop an efficient yet effective implementation of the D$^2$ST-Adapter, incorporating the specially devised anisotropic Deformable Spatio-Temporal Attention as its pivotal operation. This mechanism can be individually tailored for two pathways with anisotropic sampling densities along the spatial and temporal domains in 3D spatio-temporal space, enabling disentangled encoding of spatial and temporal features while maintaining a lightweight design. Extensive experiments by instantiating our method on both pre-trained ResNet and ViT demonstrate the superiority of our method over state-of-the-art methods. Our method is particularly well-suited to challenging scenarios where temporal dynamics are critical for action recognition. Code is available at https://github.com/qizhongtan/D2ST-Adapter. △ Less

Submitted 30 June, 2025; v1 submitted 3 December, 2023; originally announced December 2023.

Comments: Accepted by ICCV2025

Hierarchical Contrastive Learning for Pattern-Generalizable Image Corruption Detection

Authors: Xin Feng, Yifeng Xu, Guangming Lu, Wenjie Pei

Abstract: Effective image restoration with large-size corruptions, such as blind image inpainting, entails precise detection of corruption region masks which remains extremely challenging due to diverse shapes and patterns of corruptions. In this work, we present a novel method for automatic corruption detection, which allows for blind corruption restoration without known corruption masks. Specifically, we… ▽ More Effective image restoration with large-size corruptions, such as blind image inpainting, entails precise detection of corruption region masks which remains extremely challenging due to diverse shapes and patterns of corruptions. In this work, we present a novel method for automatic corruption detection, which allows for blind corruption restoration without known corruption masks. Specifically, we develop a hierarchical contrastive learning framework to detect corrupted regions by capturing the intrinsic semantic distinctions between corrupted and uncorrupted regions. In particular, our model detects the corrupted mask in a coarse-to-fine manner by first predicting a coarse mask by contrastive learning in low-resolution feature space and then refines the uncertain area of the mask by high-resolution contrastive learning. A specialized hierarchical interaction mechanism is designed to facilitate the knowledge propagation of contrastive learning in different scales, boosting the modeling performance substantially. The detected multi-scale corruption masks are then leveraged to guide the corruption restoration. Detecting corrupted regions by learning the contrastive distinctions rather than the semantic patterns of corruptions, our model has well generalization ability across different corruption patterns. Extensive experiments demonstrate following merits of our model: 1) the superior performance over other methods on both corruption detection and various image restoration tasks including blind inpainting and watermark removal, and 2) strong generalization across different corruption patterns such as graffiti, random noise or other image content. Codes and trained weights are available at https://github.com/xyfJASON/HCL . △ Less

Submitted 27 August, 2023; originally announced August 2023.

Comments: ICCV 2023

Scene-Generalizable Interactive Segmentation of Radiance Fields

Authors: Songlin Tang, Wenjie Pei, Xin Tao, Tanghui Jia, Guangming Lu, Yu-Wing Tai

Abstract: Existing methods for interactive segmentation in radiance fields entail scene-specific optimization and thus cannot generalize across different scenes, which greatly limits their applicability. In this work we make the first attempt at Scene-Generalizable Interactive Segmentation in Radiance Fields (SGISRF) and propose a novel SGISRF method, which can perform 3D object segmentation for novel (unse… ▽ More Existing methods for interactive segmentation in radiance fields entail scene-specific optimization and thus cannot generalize across different scenes, which greatly limits their applicability. In this work we make the first attempt at Scene-Generalizable Interactive Segmentation in Radiance Fields (SGISRF) and propose a novel SGISRF method, which can perform 3D object segmentation for novel (unseen) scenes represented by radiance fields, guided by only a few interactive user clicks in a given set of multi-view 2D images. In particular, the proposed SGISRF focuses on addressing three crucial challenges with three specially designed techniques. First, we devise the Cross-Dimension Guidance Propagation to encode the scarce 2D user clicks into informative 3D guidance representations. Second, the Uncertainty-Eliminated 3D Segmentation module is designed to achieve efficient yet effective 3D segmentation. Third, Concealment-Revealed Supervised Learning scheme is proposed to reveal and correct the concealed 3D segmentation errors resulted from the supervision in 2D space with only 2D mask annotations. Extensive experiments on two real-world challenging benchmarks covering diverse scenes demonstrate 1) effectiveness and scene-generalizability of the proposed method, 2) favorable performance compared to classical method requiring scene-specific optimization. △ Less

Submitted 9 August, 2023; originally announced August 2023.

Feature Decoupling-Recycling Network for Fast Interactive Segmentation

Authors: Huimin Zeng, Weinong Wang, Xin Tao, Zhiwei Xiong, Yu-Wing Tai, Wenjie Pei

Abstract: Recent interactive segmentation methods iteratively take source image, user guidance and previously predicted mask as the input without considering the invariant nature of the source image. As a result, extracting features from the source image is repeated in each interaction, resulting in substantial computational redundancy. In this work, we propose the Feature Decoupling-Recycling Network (FDRN… ▽ More Recent interactive segmentation methods iteratively take source image, user guidance and previously predicted mask as the input without considering the invariant nature of the source image. As a result, extracting features from the source image is repeated in each interaction, resulting in substantial computational redundancy. In this work, we propose the Feature Decoupling-Recycling Network (FDRN), which decouples the modeling components based on their intrinsic discrepancies and then recycles components for each user interaction. Thus, the efficiency of the whole interactive process can be significantly improved. To be specific, we apply the Decoupling-Recycling strategy from three perspectives to address three types of discrepancies, respectively. First, our model decouples the learning of source image semantics from the encoding of user guidance to process two types of input domains separately. Second, FDRN decouples high-level and low-level features from stratified semantic representations to enhance feature learning. Third, during the encoding of user guidance, current user guidance is decoupled from historical guidance to highlight the effect of current user guidance. We conduct extensive experiments on 6 datasets from different domains and modalities, which demonstrate the following merits of our model: 1) superior efficiency than other methods, particularly advantageous in challenging scenarios requiring long-term interactions (up to 4.25x faster), while achieving favorable segmentation performance; 2) strong applicability to various methods serving as a universal enhancement technique; 3) well cross-task generalizability, e.g., to medical image segmentation, and robustness against misleading user guidance. △ Less

Submitted 8 August, 2023; v1 submitted 7 August, 2023; originally announced August 2023.

Comments: Accepted to ACM MM 2023

Boosting Few-shot 3D Point Cloud Segmentation via Query-Guided Enhancement

Authors: Zhenhua Ning, Zhuotao Tian, Guangming Lu, Wenjie Pei

Abstract: Although extensive research has been conducted on 3D point cloud segmentation, effectively adapting generic models to novel categories remains a formidable challenge. This paper proposes a novel approach to improve point cloud few-shot segmentation (PC-FSS) models. Unlike existing PC-FSS methods that directly utilize categorical information from support prototypes to recognize novel classes in que… ▽ More Although extensive research has been conducted on 3D point cloud segmentation, effectively adapting generic models to novel categories remains a formidable challenge. This paper proposes a novel approach to improve point cloud few-shot segmentation (PC-FSS) models. Unlike existing PC-FSS methods that directly utilize categorical information from support prototypes to recognize novel classes in query samples, our method identifies two critical aspects that substantially enhance model performance by reducing contextual gaps between support prototypes and query features. Specifically, we (1) adapt support background prototypes to match query context while removing extraneous cues that may obscure foreground and background in query samples, and (2) holistically rectify support prototypes under the guidance of query features to emulate the latter having no semantic gap to the query targets. Our proposed designs are agnostic to the feature extractor, rendering them readily applicable to any prototype-based methods. The experimental results on S3DIS and ScanNet demonstrate notable practical benefits, as our approach achieves significant improvements while still maintaining high efficiency. The code for our approach is available at https://github.com/AaronNZH/Boosting-Few-shot-3D-Point-Cloud-Segmentation-via-Query-Guided-Enhancement △ Less

Submitted 8 August, 2023; v1 submitted 6 August, 2023; originally announced August 2023.

Comments: Accepted to ACM MM 2023

Reliability-Hierarchical Memory Network for Scribble-Supervised Video Object Segmentation

Authors: Zikun Zhou, Kaige Mao, Wenjie Pei, Hongpeng Wang, Yaowei Wang, Zhenyu He

Abstract: This paper aims to solve the video object segmentation (VOS) task in a scribble-supervised manner, in which VOS models are not only trained by the sparse scribble annotations but also initialized with the sparse target scribbles for inference. Thus, the annotation burdens for both training and initialization can be substantially lightened. The difficulties of scribble-supervised VOS lie in two asp… ▽ More This paper aims to solve the video object segmentation (VOS) task in a scribble-supervised manner, in which VOS models are not only trained by the sparse scribble annotations but also initialized with the sparse target scribbles for inference. Thus, the annotation burdens for both training and initialization can be substantially lightened. The difficulties of scribble-supervised VOS lie in two aspects. On the one hand, it requires the powerful ability to learn from the sparse scribble annotations during training. On the other hand, it demands strong reasoning capability during inference given only a sparse initial target scribble. In this work, we propose a Reliability-Hierarchical Memory Network (RHMNet) to predict the target mask in a step-wise expanding strategy w.r.t. the memory reliability level. To be specific, RHMNet first only uses the memory in the high-reliability level to locate the region with high reliability belonging to the target, which is highly similar to the initial target scribble. Then it expands the located high-reliability region to the entire target conditioned on the region itself and the memories in all reliability levels. Besides, we propose a scribble-supervised learning mechanism to facilitate the learning of our model to predict dense results. It mines the pixel-level relation within the single frame and the frame-level relation within the sequence to take full advantage of the scribble annotations in sequence training samples. The favorable performance on two popular benchmarks demonstrates that our method is promising. △ Less

Submitted 25 March, 2023; originally announced March 2023.

Comments: This project is available at https://github.com/mkg1204/RHMNet-for-SSVOS

Audio2Gestures: Generating Diverse Gestures from Audio

Authors: Jing Li, Di Kang, Wenjie Pei, Xuefei Zhe, Ying Zhang, Linchao Bao, Zhenyu He

Abstract: People may perform diverse gestures affected by various mental and physical factors when speaking the same sentences. This inherent one-to-many relationship makes co-speech gesture generation from audio particularly challenging. Conventional CNNs/RNNs assume one-to-one mapping, and thus tend to predict the average of all possible target motions, easily resulting in plain/boring motions during infe… ▽ More People may perform diverse gestures affected by various mental and physical factors when speaking the same sentences. This inherent one-to-many relationship makes co-speech gesture generation from audio particularly challenging. Conventional CNNs/RNNs assume one-to-one mapping, and thus tend to predict the average of all possible target motions, easily resulting in plain/boring motions during inference. So we propose to explicitly model the one-to-many audio-to-motion mapping by splitting the cross-modal latent code into shared code and motion-specific code. The shared code is expected to be responsible for the motion component that is more correlated to the audio while the motion-specific code is expected to capture diverse motion information that is more independent of the audio. However, splitting the latent code into two parts poses extra training difficulties. Several crucial training losses/strategies, including relaxed motion loss, bicycle constraint, and diversity loss, are designed to better train the VAE. Experiments on both 3D and 2D motion datasets verify that our method generates more realistic and diverse motions than previous state-of-the-art methods, quantitatively and qualitatively. Besides, our formulation is compatible with discrete cosine transformation (DCT) modeling and other popular backbones (\textit{i.e.} RNN, Transformer). As for motion losses and quantitative motion evaluation, we find structured losses/metrics (\textit{e.g.} STFT) that consider temporal and/or spatial context complement the most commonly used point-wise losses (\textit{e.g.} PCK), resulting in better motion dynamics and more nuanced motion details. Finally, we demonstrate that our method can be readily used to generate motion sequences with user-specified motion clips on the timeline. △ Less

Submitted 16 January, 2023; originally announced January 2023.

Comments: arXiv admin note: substantial text overlap with arXiv:2108.06720

Activating the Discriminability of Novel Classes for Few-shot Segmentation

Authors: Dianwen Mei, Wei Zhuo, Jiandong Tian, Guangming Lu, Wenjie Pei

Abstract: Despite the remarkable success of existing methods for few-shot segmentation, there remain two crucial challenges. First, the feature learning for novel classes is suppressed during the training on base classes in that the novel classes are always treated as background. Thus, the semantics of novel classes are not well learned. Second, most of existing methods fail to consider the underlying seman… ▽ More Despite the remarkable success of existing methods for few-shot segmentation, there remain two crucial challenges. First, the feature learning for novel classes is suppressed during the training on base classes in that the novel classes are always treated as background. Thus, the semantics of novel classes are not well learned. Second, most of existing methods fail to consider the underlying semantic gap between the support and the query resulting from the representative bias by the scarce support samples. To circumvent these two challenges, we propose to activate the discriminability of novel classes explicitly in both the feature encoding stage and the prediction stage for segmentation. In the feature encoding stage, we design the Semantic-Preserving Feature Learning module (SPFL) to first exploit and then retain the latent semantics contained in the whole input image, especially those in the background that belong to novel classes. In the prediction stage for segmentation, we learn an Self-Refined Online Foreground-Background classifier (SROFB), which is able to refine itself using the high-confidence pixels of query image to facilitate its adaptation to the query image and bridge the support-query semantic gap. Extensive experiments on PASCAL-5$^i$ and COCO-20$^i$ datasets demonstrates the advantages of these two novel designs both quantitatively and qualitatively. △ Less

Submitted 2 December, 2022; originally announced December 2022.

Explaining Deep Convolutional Neural Networks for Image Classification by Evolving Local Interpretable Model-agnostic Explanations

Authors: Bin Wang, Wenbin Pei, Bing Xue, Mengjie Zhang

Abstract: Deep convolutional neural networks have proven their effectiveness, and have been acknowledged as the most dominant method for image classification. However, a severe drawback of deep convolutional neural networks is poor explainability. Unfortunately, in many real-world applications, users need to understand the rationale behind the predictions of deep convolutional neural networks when determini… ▽ More Deep convolutional neural networks have proven their effectiveness, and have been acknowledged as the most dominant method for image classification. However, a severe drawback of deep convolutional neural networks is poor explainability. Unfortunately, in many real-world applications, users need to understand the rationale behind the predictions of deep convolutional neural networks when determining whether they should trust the predictions or not. To resolve this issue, a novel genetic algorithm-based method is proposed for the first time to automatically evolve local explanations that can assist users to assess the rationality of the predictions. Furthermore, the proposed method is model-agnostic, i.e., it can be utilised to explain any deep convolutional neural network models. In the experiments, ResNet is used as an example model to be explained, and the ImageNet dataset is selected as the benchmark dataset. DenseNet and MobileNet are further explained to demonstrate the model-agnostic characteristic of the proposed method. The evolved local explanations on four images, randomly selected from ImageNet, are presented, which show that the evolved local explanations are straightforward to be recognised by humans. Moreover, the evolved explanations can explain the predictions of deep convolutional neural networks on all four images very well by successfully capturing meaningful interpretable features of the sample images. Further analysis based on the 30 runs of the experiments exhibits that the evolved local explanations can also improve the probabilities/confidences of the deep convolutional neural network models in making the predictions. The proposed method can obtain local explanations within one minute, which is more than ten times faster than LIME (the state-of-the-art method). △ Less

Submitted 25 March, 2025; v1 submitted 28 November, 2022; originally announced November 2022.

Semantic-Aware Local-Global Vision Transformer

Authors: Jiatong Zhang, Zengwei Yao, Fanglin Chen, Guangming Lu, Wenjie Pei

Abstract: Vision Transformers have achieved remarkable progresses, among which Swin Transformer has demonstrated the tremendous potential of Transformer for vision tasks. It surmounts the key challenge of high computational complexity by performing local self-attention within shifted windows. In this work we propose the Semantic-Aware Local-Global Vision Transformer (SALG), to further investigate two potent… ▽ More Vision Transformers have achieved remarkable progresses, among which Swin Transformer has demonstrated the tremendous potential of Transformer for vision tasks. It surmounts the key challenge of high computational complexity by performing local self-attention within shifted windows. In this work we propose the Semantic-Aware Local-Global Vision Transformer (SALG), to further investigate two potential improvements towards Swin Transformer. First, unlike Swin Transformer that performs uniform partition to produce equal size of regular windows for local self-attention, our SALG performs semantic segmentation in an unsupervised way to explore the underlying semantic priors in the image. As a result, each segmented region can correspond to a semantically meaningful part in the image, potentially leading to more effective features within each of segmented regions. Second, instead of only performing local self-attention within local windows as Swin Transformer does, the proposed SALG performs both 1) local intra-region self-attention for learning fine-grained features within each region and 2) global inter-region feature propagation for modeling global dependencies among all regions. Consequently, our model is able to obtain the global view when learning features for each token, which is the essential advantage of Transformer. Owing to the explicit modeling of the semantic priors and the proposed local-global modeling mechanism, our SALG is particularly advantageous for small-scale models when the modeling capacity is not sufficient for other models to learn semantics implicitly. Extensive experiments across various vision tasks demonstrates the merit of our model over other vision Transformers, especially in the small-scale modeling scenarios. △ Less

Submitted 26 November, 2022; originally announced November 2022.

Alleviating the Sample Selection Bias in Few-shot Learning by Removing Projection to the Centroid

Authors: Jing Xu, Xu Luo, Xinglin Pan, Wenjie Pei, Yanan Li, Zenglin Xu

Abstract: Few-shot learning (FSL) targets at generalization of vision models towards unseen tasks without sufficient annotations. Despite the emergence of a number of few-shot learning methods, the sample selection bias problem, i.e., the sensitivity to the limited amount of support data, has not been well understood. In this paper, we find that this problem usually occurs when the positions of support samp… ▽ More Few-shot learning (FSL) targets at generalization of vision models towards unseen tasks without sufficient annotations. Despite the emergence of a number of few-shot learning methods, the sample selection bias problem, i.e., the sensitivity to the limited amount of support data, has not been well understood. In this paper, we find that this problem usually occurs when the positions of support samples are in the vicinity of task centroid -- the mean of all class centroids in the task. This motivates us to propose an extremely simple feature transformation to alleviate this problem, dubbed Task Centroid Projection Removing (TCPR). TCPR is applied directly to all image features in a given task, aiming at removing the dimension of features along the direction of the task centroid. While the exact task centroid cannot be accurately obtained from limited data, we estimate it using base features that are each similar to one of the support features. Our method effectively prevents features from being too close to the task centroid. Extensive experiments over ten datasets from different domains show that TCPR can reliably improve classification accuracy across various feature extractors, training algorithms and datasets. The code has been made available at https://github.com/KikimorMay/FSL-TCBR. △ Less

Submitted 30 October, 2022; originally announced October 2022.

Comments: Accepted at NeurIPS 2022

SSORN: Self-Supervised Outlier Removal Network for Robust Homography Estimation

Authors: Yi Li, Wenjie Pei, Zhenyu He

Abstract: The traditional homography estimation pipeline consists of four main steps: feature detection, feature matching, outlier removal and transformation estimation. Recent deep learning models intend to address the homography estimation problem using a single convolutional network. While these models are trained in an end-to-end fashion to simplify the homography estimation problem, they lack the featu… ▽ More The traditional homography estimation pipeline consists of four main steps: feature detection, feature matching, outlier removal and transformation estimation. Recent deep learning models intend to address the homography estimation problem using a single convolutional network. While these models are trained in an end-to-end fashion to simplify the homography estimation problem, they lack the feature matching step and/or the outlier removal step, which are important steps in the traditional homography estimation pipeline. In this paper, we attempt to build a deep learning model that mimics all four steps in the traditional homography estimation pipeline. In particular, the feature matching step is implemented using the cost volume technique. To remove outliers in the cost volume, we treat this outlier removal problem as a denoising problem and propose a novel self-supervised loss to solve the problem. Extensive experiments on synthetic and real datasets demonstrate that the proposed model outperforms existing deep learning models. △ Less

Submitted 30 August, 2022; originally announced August 2022.

Layout-Bridging Text-to-Image Synthesis

Authors: Jiadong Liang, Wenjie Pei, Feng Lu

Abstract: The crux of text-to-image synthesis stems from the difficulty of preserving the cross-modality semantic consistency between the input text and the synthesized image. Typical methods, which seek to model the text-to-image mapping directly, could only capture keywords in the text that indicates common objects or actions but fail to learn their spatial distribution patterns. An effective way to circu… ▽ More The crux of text-to-image synthesis stems from the difficulty of preserving the cross-modality semantic consistency between the input text and the synthesized image. Typical methods, which seek to model the text-to-image mapping directly, could only capture keywords in the text that indicates common objects or actions but fail to learn their spatial distribution patterns. An effective way to circumvent this limitation is to generate an image layout as guidance, which is attempted by a few methods. Nevertheless, these methods fail to generate practically effective layouts due to the diversity of input text and object location. In this paper we push for effective modeling in both text-to-layout generation and layout-to-image synthesis. Specifically, we formulate the text-to-layout generation as a sequence-to-sequence modeling task, and build our model upon Transformer to learn the spatial relationships between objects by modeling the sequential dependencies between them. In the stage of layout-to-image synthesis, we focus on learning the textual-visual semantic alignment per object in the layout to precisely incorporate the input text into the layout-to-image synthesizing process. To evaluate the quality of generated layout, we design a new metric specifically, dubbed Layout Quality Score, which considers both the absolute distribution errors of bounding boxes in the layout and the mutual spatial relationships between them. Extensive experiments on three datasets demonstrate the superior performance of our method over state-of-the-art methods on both predicting the layout and synthesizing the image from the given text. △ Less

Submitted 12 August, 2022; originally announced August 2022.

Learning Generalizable Latent Representations for Novel Degradations in Super Resolution

Authors: Fengjun Li, Xin Feng, Fanglin Chen, Guangming Lu, Wenjie Pei

Abstract: Typical methods for blind image super-resolution (SR) focus on dealing with unknown degradations by directly estimating them or learning the degradation representations in a latent space. A potential limitation of these methods is that they assume the unknown degradations can be simulated by the integration of various handcrafted degradations (e.g., bicubic downsampling), which is not necessarily… ▽ More Typical methods for blind image super-resolution (SR) focus on dealing with unknown degradations by directly estimating them or learning the degradation representations in a latent space. A potential limitation of these methods is that they assume the unknown degradations can be simulated by the integration of various handcrafted degradations (e.g., bicubic downsampling), which is not necessarily true. The real-world degradations can be beyond the simulation scope by the handcrafted degradations, which are referred to as novel degradations. In this work, we propose to learn a latent representation space for degradations, which can be generalized from handcrafted (base) degradations to novel degradations. The obtained representations for a novel degradation in this latent space are then leveraged to generate degraded images consistent with the novel degradation to compose paired training data for SR model. Furthermore, we perform variational inference to match the posterior of degradations in latent representation space with a prior distribution (e.g., Gaussian distribution). Consequently, we are able to sample more high-quality representations for a novel degradation to augment the training data for SR model. We conduct extensive experiments on both synthetic and real-world datasets to validate the effectiveness and advantages of our method for blind super-resolution with novel degradations. △ Less

Submitted 25 July, 2022; originally announced July 2022.

Few-Shot Object Detection by Knowledge Distillation Using Bag-of-Visual-Words Representations

Authors: Wenjie Pei, Shuang Wu, Dianwen Mei, Fanglin Chen, Jiandong Tian, Guangming Lu

Abstract: While fine-tuning based methods for few-shot object detection have achieved remarkable progress, a crucial challenge that has not been addressed well is the potential class-specific overfitting on base classes and sample-specific overfitting on novel classes. In this work we design a novel knowledge distillation framework to guide the learning of the object detector and thereby restrain the overfi… ▽ More While fine-tuning based methods for few-shot object detection have achieved remarkable progress, a crucial challenge that has not been addressed well is the potential class-specific overfitting on base classes and sample-specific overfitting on novel classes. In this work we design a novel knowledge distillation framework to guide the learning of the object detector and thereby restrain the overfitting in both the pre-training stage on base classes and fine-tuning stage on novel classes. To be specific, we first present a novel Position-Aware Bag-of-Visual-Words model for learning a representative bag of visual words (BoVW) from a limited size of image set, which is used to encode general images based on the similarities between the learned visual words and an image. Then we perform knowledge distillation based on the fact that an image should have consistent BoVW representations in two different feature spaces. To this end, we pre-learn a feature space independently from the object detection, and encode images using BoVW in this space. The obtained BoVW representation for an image can be considered as distilled knowledge to guide the learning of object detector: the extracted features by the object detector for the same image are expected to derive the consistent BoVW representations with the distilled knowledge. Extensive experiments validate the effectiveness of our method and demonstrate the superiority over other state-of-the-art methods. △ Less

Submitted 25 July, 2022; originally announced July 2022.

Self-Support Few-Shot Semantic Segmentation

Authors: Qi Fan, Wenjie Pei, Yu-Wing Tai, Chi-Keung Tang

Abstract: Existing few-shot segmentation methods have achieved great progress based on the support-query matching framework. But they still heavily suffer from the limited coverage of intra-class variations from the few-shot supports provided. Motivated by the simple Gestalt principle that pixels belonging to the same object are more similar than those to different objects of same class, we propose a novel… ▽ More Existing few-shot segmentation methods have achieved great progress based on the support-query matching framework. But they still heavily suffer from the limited coverage of intra-class variations from the few-shot supports provided. Motivated by the simple Gestalt principle that pixels belonging to the same object are more similar than those to different objects of same class, we propose a novel self-support matching strategy to alleviate this problem, which uses query prototypes to match query features, where the query prototypes are collected from high-confidence query predictions. This strategy can effectively capture the consistent underlying characteristics of the query objects, and thus fittingly match query features. We also propose an adaptive self-support background prototype generation module and self-support loss to further facilitate the self-support matching procedure. Our self-support network substantially improves the prototype quality, benefits more improvement from stronger backbones and more supports, and achieves SOTA on multiple datasets. Codes are at \url{https://github.com/fanq15/SSP}. △ Less

Submitted 23 July, 2022; originally announced July 2022.

Comments: ECCV 2022

Multi-Faceted Distillation of Base-Novel Commonality for Few-shot Object Detection

Authors: Shuang Wu, Wenjie Pei, Dianwen Mei, Fanglin Chen, Jiandong Tian, Guangming Lu

Abstract: Most of existing methods for few-shot object detection follow the fine-tuning paradigm, which potentially assumes that the class-agnostic generalizable knowledge can be learned and transferred implicitly from base classes with abundant samples to novel classes with limited samples via such a two-stage training strategy. However, it is not necessarily true since the object detector can hardly disti… ▽ More Most of existing methods for few-shot object detection follow the fine-tuning paradigm, which potentially assumes that the class-agnostic generalizable knowledge can be learned and transferred implicitly from base classes with abundant samples to novel classes with limited samples via such a two-stage training strategy. However, it is not necessarily true since the object detector can hardly distinguish between class-agnostic knowledge and class-specific knowledge automatically without explicit modeling. In this work we propose to learn three types of class-agnostic commonalities between base and novel classes explicitly: recognition-related semantic commonalities, localization-related semantic commonalities and distribution commonalities. We design a unified distillation framework based on a memory bank, which is able to perform distillation of all three types of commonalities jointly and efficiently. Extensive experiments demonstrate that our method can be readily integrated into most of existing fine-tuning based methods and consistently improve the performance by a large margin. △ Less

Submitted 3 November, 2022; v1 submitted 22 July, 2022; originally announced July 2022.

Comments: Accepted to ECCV 2022