arXiv subjects

Jian Zhang

Publications and source records attributed to Jian Zhang.

At least 235 records · Page 13Linked to original sources

arXiv2025

Write Your Own CodeChecker: An Automated Test-Driven Checker Development Approach with LLMs

With the rising demand for code quality assurance, developers are not only utilizing existing static code checkers but also seeking custom checkers to satisfy their specific needs. Nowadays, various code-checking frameworks provide extensive checker customization interfaces to meet this need. However, both the abstract checking logic and the complex API usage of large-scale checker frameworks make this task challenging. To this end, automated code checker generation is anticipated to ease the burden of checker development. In this paper, we propose AutoChecker, an innovative LLM-powered approach that can write code checkers automatically based on only a rule description and a test suite. To achieve comprehensive checking logic, AutoChecker incrementally updates the checker's logic by focusing on solving one selected case each time. To obtain precise API knowledge, during each iteration, it leverages fine-grained logic-guided API-context retrieval, where it first decomposes the checking logic into a series of sub-operations and then retrieves checker-related API-contexts for each sub-operation. For evaluation, we apply AutoChecker, five baselines, and three ablation methods using multiple LLMs to generate checkers for 20 randomly selected PMD rules. Experimental results show that AutoChecker significantly outperforms others across all effectiveness metrics, with an average test pass rate of 82.28%. Additionally, the checkers generated by AutoChecker can be successfully applied to real-world projects, matching the performance of official checkers.

cs.SE↗

arXiv2025

Unsupervised Part Discovery via Descriptor-Based Masked Image Restoration with Optimized Constraints

Part-level features are crucial for image understanding, but few studies focus on them because of the lack of fine-grained labels. Although unsupervised part discovery can eliminate the reliance on labels, most of them cannot maintain robustness across various categories and scenarios, which restricts their application range. To overcome this limitation, we present a more effective paradigm for unsupervised part discovery, named Masked Part Autoencoder (MPAE). It first learns part descriptors as well as a feature map from the inputs and produces patch features from a masked version of the original images. Then, the masked regions are filled with the learned part descriptors based on the similarity between the local features and descriptors. By restoring these masked patches using the part descriptors, they become better aligned with their part shapes, guided by appearance features from unmasked patches. Finally, MPAE robustly discovers meaningful parts that closely match the actual object shapes, even in complex scenarios. Moreover, several looser yet more effective constraints are proposed to enable MPAE to identify the presence of parts across various scenarios and categories in an unsupervised manner. This provides the foundation for addressing challenges posed by occlusion and for exploring part similarity across multiple categories. Extensive experiments demonstrate that our method robustly discovers meaningful parts across various categories and scenarios. The code is available at the project https://github.com/Jiahao-UTS/MPAE.

cs.CV↗

arXiv2025

Weak low-temperature ferromagnetism and linear magnetoresistance in Lu$_{0.75}$Fe$_6$Sn$_6$ with a disordered HfFe$_6$Ge$_6$-type structure

We report the synthesis of Lu$_{0.75}$Fe$_6$Sn$_6$ single crystals with a Fe-kagome lattice using a self-flux method. The crystal structure, magnetic, thermodynamic and electrical transport properties were investigated. Structure refinement reveals that Lu$_{0.75}$Fe$_6$Sn$_6$ has a HfFe$_6$Ge$_6$-type structure as the major framework intergrown with a CoSn-type structure, leading to a vacancy of 25% on the Lu-site and disorder on the Sn-site. It exhibits a significant magnetic anisotropy with weak ferromagnetism in the ab-plane below 40 K and antiferromagnetic behavior along the c-axis. The weak ferromagnetism is due to the canted antiferromagnetism with magnetic moment deviating from the $c$-axis to the ab-plane. Besides, an anisotropic non-saturated linear magnetoresistance is also observed in Lu$_{0.75}$Fe$_6$Sn$_6$, probably resulting from the structural disorder in the sample.

cond-mat.mtrl-sci↗

arXiv2025

PointGAC: Geometric-Aware Codebook for Masked Point Cloud Modeling

Most masked point cloud modeling (MPM) methods follow a regression paradigm to reconstruct the coordinate or feature of masked regions. However, they tend to over-constrain the model to learn the details of the masked region, resulting in failure to capture generalized features. To address this limitation, we propose \textbf{\textit{PointGAC}}, a novel clustering-based MPM method that aims to align the feature distribution of masked regions. Specially, it features an online codebook-guided teacher-student framework. Firstly, it presents a geometry-aware partitioning strategy to extract initial patches. Then, the teacher model updates a codebook via online k-means based on features extracted from the complete patches. This procedure facilitates codebook vectors to become cluster centers. Afterward, we assigns the unmasked features to their corresponding cluster centers, and the student model aligns the assignment for the reconstructed masked features. This strategy focuses on identifying the cluster centers to which the masked features belong, enabling the model to learn more generalized feature representations. Benefiting from a proposed codebook maintenance mechanism, codebook vectors are actively updated, which further increases the efficiency of semantic feature learning. Experiments validate the effectiveness of the proposed method on various downstream tasks. Code is available at https://github.com/LAB123-tech/PointGAC

cs.CV↗

arXiv2025

TalkFashion: Intelligent Virtual Try-On Assistant Based on Multimodal Large Language Model

Virtual try-on has made significant progress in recent years. This paper addresses how to achieve multifunctional virtual try-on guided solely by text instructions, including full outfit change and local editing. Previous methods primarily relied on end-to-end networks to perform single try-on tasks, lacking versatility and flexibility. We propose TalkFashion, an intelligent try-on assistant that leverages the powerful comprehension capabilities of large language models to analyze user instructions and determine which task to execute, thereby activating different processing pipelines accordingly. Additionally, we introduce an instruction-based local repainting model that eliminates the need for users to manually provide masks. With the help of multi-modal models, this approach achieves fully automated local editings, enhancing the flexibility of editing tasks. The experimental results demonstrate better semantic consistency and visual quality compared to the current methods.

cs.CV↗

arXiv2025

InstantSplat: Sparse-view Gaussian Splatting in Seconds

While neural 3D reconstruction has advanced substantially, its performance significantly degrades with sparse-view data, which limits its broader applicability, since SfM is often unreliable in sparse-view scenarios where feature matches are scarce. In this paper, we introduce InstantSplat, a novel approach for addressing sparse-view 3D scene reconstruction at lightning-fast speed. InstantSplat employs a self-supervised framework that optimizes 3D scene representation and camera poses by unprojecting 2D pixels into 3D space and aligning them using differentiable neural rendering. The optimization process is initialized with a large-scale trained geometric foundation model, which provides dense priors that yield initial points through model inference, after which we further optimize all scene parameters using photometric errors. To mitigate redundancy introduced by the prior model, we propose a co-visibility-based geometry initialization, and a Gaussian-based bundle adjustment is employed to rapidly adapt both the scene representation and camera parameters without relying on a complex adaptive density control process. Overall, InstantSplat is compatible with multiple point-based representations for view synthesis and surface reconstruction. It achieves an acceleration of over 30x in reconstruction and improves visual quality (SSIM) from 0.3755 to 0.7624 compared to traditional SfM with 3D-GS.

cs.CV↗

arXiv2025

SWinMamba: Serpentine Window State Space Model for Vascular Segmentation

Vascular segmentation in medical images is crucial for disease diagnosis and surgical navigation. However, the segmented vascular structure is often discontinuous due to its slender nature and inadequate prior modeling. In this paper, we propose a novel Serpentine Window Mamba (SWinMamba) to achieve accurate vascular segmentation. The proposed SWinMamba innovatively models the continuity of slender vascular structures by incorporating serpentine window sequences into bidirectional state space models. The serpentine window sequences enable efficient feature capturing by adaptively guiding global visual context modeling to the vascular structure. Specifically, the Serpentine Window Tokenizer (SWToken) adaptively splits the input image using overlapping serpentine window sequences, enabling flexible receptive fields (RFs) for vascular structure modeling. The Bidirectional Aggregation Module (BAM) integrates coherent local features in the RFs for vascular continuity representation. In addition, dual-domain learning with Spatial-Frequency Fusion Unit (SFFU) is designed to enhance the feature representation of vascular structure. Extensive experiments on three challenging datasets demonstrate that the proposed SWinMamba achieves superior performance with complete and connected vessels.

eess.IV↗

arXiv2025

Repair Ingredients Are All You Need: Improving Large Language Model-Based Program Repair via Repair Ingredients Search

Automated Program Repair (APR) techniques aim to automatically fix buggy programs. Among these, Large Language Model-based (LLM-based) approaches have shown great promise. Recent advances demonstrate that directly leveraging LLMs can achieve leading results. However, these techniques remain suboptimal in generating contextually relevant and accurate patches, as they often overlook repair ingredients crucial for practical program repair. In this paper, we propose ReinFix, a novel framework that enables LLMs to autonomously search for repair ingredients throughout both the reasoning and solution phases of bug fixing. In the reasoning phase, ReinFix integrates static analysis tools to retrieve internal ingredients, such as variable definitions, to assist the LLM in root cause analysis when it encounters difficulty understanding the context. During the solution phase, when the LLM lacks experience in fixing specific bugs, ReinFix searches for external ingredients from historical bug fixes with similar bug patterns, leveraging both the buggy code and its root cause to guide the LLM in identifying appropriate repair actions, thereby increasing the likelihood of generating correct patches. Evaluations on two popular benchmarks (Defects4J V1.2 and V2.0) demonstrate the effectiveness of our approach over SOTA baselines. Notably, ReinFix fixes 146 bugs, which is 32 more than the baselines on Defects4J V1.2. On Defects4J V2.0, ReinFix fixes 38 more bugs than the SOTA. Importantly, when evaluating on the recent benchmarks that are free of data leakage risk, ReinFix also maintains the best performance.

cs.SE↗

arXiv2025

VFArchē: A Dual-Mode Framework for Locating Vulnerable Functions in Open-Source Software

Software Composition Analysis (SCA) has become pivotal in addressing vulnerabilities inherent in software project dependencies. In particular, reachability analysis is increasingly used in Open-Source Software (OSS) projects to identify reachable vulnerabilities (e.g., CVEs) through call graphs, enabling a focus on exploitable risks. Performing reachability analysis typically requires the vulnerable function (VF) to track the call chains from downstream applications. However, such crucial information is usually unavailable in modern vulnerability databases like NVD. While directly extracting VF from modified functions in vulnerability patches is intuitive, patches are not always available. Moreover, our preliminary study shows that over 26% of VF do not exist in the modified functions. Meanwhile, simply ignoring patches to search vulnerable functions suffers from overwhelming noises and lexical gaps between descriptions and source code. Given that almost half of the vulnerabilities are equipped with patches, a holistic solution that handles both scenarios with and without patches is required. To meet real-world needs and automatically localize VF, we present VFArchē, a dual-mode approach designed for disclosed vulnerabilities, applicable in scenarios with or without available patch links. The experimental results of VFArchē on our constructed benchmark dataset demonstrate significant efficacy regarding three metrics, achieving 1.3x and 1.9x Mean Reciprocal Rank over the best baselines for Patch-present and Patch-absent modes, respectively. Moreover, VFArchē has proven its applicability in real-world scenarios by successfully locating VF for 43 out of 50 latest vulnerabilities with reasonable efforts and significantly reducing 78-89% false positives of SCA tools.

cs.SE↗

arXiv2025

VQ-Insight: Teaching VLMs for AI-Generated Video Quality Understanding via Progressive Visual Reinforcement Learning

Recent advances in AI-generated content (AIGC) have led to the emergence of powerful text-to-video generation models. Despite these successes, evaluating the quality of AIGC-generated videos remains challenging due to limited generalization, lack of temporal awareness, heavy reliance on large-scale annotated datasets, and the lack of effective interaction with generation models. Most current approaches rely on supervised finetuning of vision-language models (VLMs), which often require large-scale annotated datasets and tend to decouple understanding and generation. To address these shortcomings, we propose VQ-Insight, a novel reasoning-style VLM framework for AIGC video quality assessment. Our approach features: (1) a progressive video quality learning scheme that combines image quality warm-up, general task-specific temporal learning, and joint optimization with the video generation model; (2) the design of multi-dimension scoring rewards, preference comparison rewards, and temporal modeling rewards to enhance both generalization and specialization in video quality evaluation. Extensive experiments demonstrate that VQ-Insight consistently outperforms state-of-the-art baselines in preference comparison, multi-dimension scoring, and natural video scoring, bringing significant improvements for video generation tasks.

cs.CV↗

arXiv2025

Quantum Berezinian for quantum affine superalgebra $U_q(\widehat{gl}_{M|N})$

We introduce the quantum Berezinian for the quantum affine superalgebra $\mathrm{U}_q(\widehat{\mathfrak{gl}}_{M|N})$ and show that the coefficients of the quantum Berezinian belong to the center of $\mathrm{U}_q(\widehat{\gl}_{M|N})$. We also construct another family of central elements which can be expressed in the quantum Berezinian by a Liouville-type theorem. Moreover, we prove analogues of the Jacobi identities, the Schur complementary theorem, the Sylvester theorem and the MacMahon Master theorem for the generator matrices of $\mathrm{U}_q(\widehat{\gl}_{M|N})$.

math.QA↗

arXiv2025

Label-guided Facial Retouching Reversion

With the popularity of social media platforms and retouching tools, more people are beautifying their facial photos, posing challenges for fields requiring photo authenticity. To address this issue, some work has proposed makeup removal methods, but they cannot revert images involving geometric deformations caused by retouching. To tackle the problem of facial retouching reversion, we propose a framework, dubbed Re-Face, which consists of three components: a facial retouching detector, an image reversion model named FaceR, and a color correction module called Hierarchical Adaptive Instance Normalization (H-AdaIN). FaceR can utilize labels generated by the facial retouching detector as guidance to revert the retouched facial images. Then, color correction is performed using H-AdaIN to address the issue of color shift. Extensive experiments demonstrate the effectiveness of our framework and each module.

cs.CV↗

arXiv2025

CNN-Derived Elemental Abundances of LAMOST DR10 Giants: Implications for Galactic Substructures

Stellar parameters and abundances provide crucial insights into stellar and Galactic evolution studies. In this work, we developed a convolutional neural network (CNN) to estimate stellar parameters: effective temperature ($T_{\text{eff}}$), surface gravity (log $g$) and metallicity (both [Fe/H] and [M/H]) as well as six $α$-elements (C, N, O, Mg, Si, Ca) and [$α$/M]. We selected giant stars with $ 3500 \, \text{K} < T_{\text{eff}} < 5500 \, \text{K} $ and $ 0 \, \text{dex} < \log g < 3.6 \, \text{dex} $ from the LAMOST and APOGEE surveys, while requiring (S/N)$_g$ of the LAMOST low-resolution spectra $>$ 10, which leaves 1,100,858 giant stars. The spectral from LAMOST and the labels from APOGEE for 62,511 common stars were used as our training set. The corresponding test set yields scatters 50 K, 0.06 dex and 0.13 dex for $T_{\text{eff}}$, [Fe/H] and log $g$, respectively. For $α$ elements O, Mg, Si and Ca, the scatters are 0.05 dex, 0.04 dex, 0.03 and 0.04 dex, respectively. For C and N elements, the scatters are 0.07 dex and 0.05 dex. For [$α$/M] and [M/H], the scatters are 0.03 dex and 0.06 dex. The mean absolute error (MAE) of most elements are between 0.02 $-$ 0.04 dex. The predicted abundances were cross-matched with previously identified substructures PG1 and PG2, with their origins subsequently analyzed. Finally, the catalog is available at https://nadc.china-vo.org/res/r101529/.

astro-ph.GA↗

arXiv2025

Seeing is Fixing: Cross-Modal Reasoning with Multimodal LLMs for Visual Software Issue Fixing

Large language model-(LLM) based automated program repair (APR) techniques have shown promising results in resolving real-world GitHub issue tasks. Existing APR systems are primarily evaluated in unimodal settings (e.g., SWE-bench). However, these autonomous systems struggle to resolve multimodal problem scenarios (e.g., SWE-bench M) due to limitations in interpreting and leveraging visual information. In multimodal scenarios, LLMs need to rely on visual information in the graphical user interface (GUI) to understand bugs and generate fixes. To bridge this gap, we propose GUIRepair, a cross-modal reasoning approach for resolving multimodal issue scenarios by understanding and capturing visual information. Specifically, GUIRepair integrates two key components, Image2Code and Code2Image, to enhance fault comprehension and patch validation. Image2Code extracts relevant project documents based on the issue report, then applies this domain knowledge to generate the reproduced code responsible for the visual symptoms, effectively translating GUI images into executable context for better fault comprehension. Code2Image replays the visual issue scenario using the reproduced code and captures GUI renderings of the patched program to assess whether the fix visually resolves the issue, providing feedback for patch validation. We evaluate GUIRepair on SWE-bench M, and the approach demonstrates significant effectiveness. When utilizing GPT-4o as the base model, GUIRepair solves 157 instances, outperforming the best open-source baseline by 26 instances. Furthermore, when using o4-mini as the base model, GUIRepair can achieve even better results and solve 175 instances, outperforming the top commercial system by 22 instances. This emphasizes the success of our new perspective on incorporating cross-modal reasoning by understanding and capturing visual information to resolve multimodal issues.

cs.SE↗

arXiv2025

Quantitative Analysis of Performance Drop in DeepSeek Model Quantization

Recently, there is a high demand for deploying DeepSeek-R1 and V3 locally, possibly because the official service often suffers from being busy and some organizations have data privacy concerns. While single-machine deployment offers infrastructure simplicity, the models' 671B FP8 parameter configuration exceeds the practical memory limits of a standard 8-GPU machine. Quantization is a widely used technique that helps reduce model memory consumption. However, it is unclear what the performance of DeepSeek-R1 and V3 will be after being quantized. This technical report presents the first quantitative evaluation of multi-bitwidth quantization across the complete DeepSeek model spectrum. Key findings reveal that 4-bit quantization maintains little performance degradation versus FP8 while enabling single-machine deployment on standard NVIDIA GPU devices. We further propose DQ3_K_M, a dynamic 3-bit quantization method that significantly outperforms traditional Q3_K_M variant on various benchmarks, which is also comparable with 4-bit quantization (Q4_K_M) approach in most tasks. Moreover, DQ3_K_M supports single-machine deployment configurations for both NVIDIA H100/A100 and Huawei 910B. Our implementation of DQ3\_K\_M is released at https://github.com/UnicomAI/DeepSeek-Eval, containing optimized 3-bit quantized variants of both DeepSeek-R1 and DeepSeek-V3.

cs.LG↗

arXiv2025

Q-Ponder: A Unified Training Pipeline for Reasoning-based Visual Quality Assessment

Recent studies demonstrate that multimodal large language models (MLLMs) can proficiently evaluate visual quality through interpretable assessments. However, existing approaches typically treat quality scoring and reasoning descriptions as separate tasks with disjoint optimization objectives, leading to a trade-off: models adept at quality reasoning descriptions struggle with precise score regression, while score-focused models lack interpretability. This limitation hinders the full potential of MLLMs in visual quality assessment, where accuracy and interpretability should be mutually reinforcing. To address this, we propose a unified two-stage training framework comprising a cold-start stage and a reinforcement learning-based fine-tuning stage. Specifically, in the first stage, we distill high-quality data from a teacher model through expert-designed prompts, initializing reasoning capabilities via cross-entropy loss supervision. In the second stage, we introduce a novel reward with Group Relative Policy Optimization (GRPO) to jointly optimize scoring accuracy and reasoning consistency. We designate the models derived from these two stages as Q-Ponder-CI and Q-Ponder. Extensive experiments show that Q-Ponder achieves state-of-the-art (SOTA) performance on quality score regression benchmarks, delivering up to 6.5% higher SRCC on cross-domain datasets. Furthermore, Q-Ponder significantly outperforms description-based SOTA models, including its teacher model Qwen-2.5-VL-72B, particularly in description accuracy and reasonableness, demonstrating the generalization potential over diverse tasks.

cs.CV↗

arXiv2025

Fully-Geometric Cross-Attention for Point Cloud Registration

Point cloud registration approaches often fail when the overlap between point clouds is low due to noisy point correspondences. This work introduces a novel cross-attention mechanism tailored for Transformer-based architectures that tackles this problem, by fusing information from coordinates and features at the super-point level between point clouds. This formulation has remained unexplored primarily because it must guarantee rotation and translation invariance since point clouds reside in different and independent reference frames. We integrate the Gromov-Wasserstein distance into the cross-attention formulation to jointly compute distances between points across different point clouds and account for their geometric structure. By doing so, points from two distinct point clouds can attend to each other under arbitrary rigid transformations. At the point level, we also devise a self-attention mechanism that aggregates the local geometric structure information into point features for fine matching. Our formulation boosts the number of inlier correspondences, thereby yielding more precise registration results compared to state-of-the-art approaches. We have conducted an extensive evaluation on 3DMatch, 3DLoMatch, KITTI, and 3DCSR datasets.

cs.CV↗

arXiv2025

Pts3D-LLM: Studying the Impact of Token Structure for 3D Scene Understanding With Large Language Models

Effectively representing 3D scenes for Multimodal Large Language Models (MLLMs) is crucial yet challenging. Existing approaches commonly only rely on 2D image features and use varied tokenization approaches. This work presents a rigorous study of 3D token structures, systematically comparing video-based and point-based representations while maintaining consistent model backbones and parameters. We propose a novel approach that enriches visual tokens by incorporating 3D point cloud features from a Sonata pretrained Point Transformer V3 encoder. Our experiments demonstrate that merging explicit 3D features significantly boosts performance. Furthermore, we show that point-based token structures can rival video-based ones when the points are cleverly sampled and ordered. Our best models from both structures achieve state-of-the-art results on multiple 3D understanding benchmarks. We emphasize our analysis of token structures as a key contribution, alongside transparent reporting of results averaged over multiple seeds, a practice we believe is vital for robust progress in the field.

cs.CV↗