arXiv subjects

Jun Du

Publications and source records attributed to Jun Du.

At least 37 records · Page 2Linked to original sources

arXiv2026

WebTestBench: Evaluating Computer-Use Agents towards End-to-End Automated Web Testing

The emergence of Large Language Models (LLMs) has catalyzed a paradigm shift in programming, giving rise to "vibe coding", where users can build complete projects and even control computers using natural language instructions. This paradigm has driven automated webpage development, but it introduces a new requirement about how to automatically verify whether the web functionalities are reliably implemented. Existing works struggle to adapt, relying on static visual similarity or predefined checklists that constrain their utility in open-ended environments. Furthermore, they overlook a vital aspect of software quality, namely latent logical constraints. To address these gaps, we introduce WebTestBench, a benchmark for evaluating end-to-end automated web testing. WebTestBench encompasses comprehensive dimensions across diverse web application categories. We decompose the testing process into two cascaded sub-tasks, checklist generation and defect detection, and propose WebTester, a baseline framework for this task. Evaluating popular LLMs with WebTester reveals severe challenges, including insufficient test completeness, detection bottlenecks, and long-horizon interaction unreliability. These findings expose a substantial gap between current computer-use agent capabilities and industrial-grade deployment demands. We hope that WebTestBench provides valuable insights and guidance for advancing end-to-end automated web testing. Our dataset and code are available at https://github.com/friedrichor/WebTestBench.

cs.SE↗

arXiv2026

TDATR: Improving End-to-End Table Recognition via Table Detail-Aware Learning and Cell-Level Visual Alignment

Tables are pervasive in diverse documents, making table recognition (TR) a fundamental task in document analysis. Existing modular TR pipelines separately model table structure and content, leading to suboptimal integration and complex workflows. End-to-end approaches rely heavily on large-scale TR data and struggle in data-constrained scenarios. To address these issues, we propose TDATR (Table Detail-Aware Table Recognition) improves end-to-end TR through table detail-aware learning and cell-level visual alignment. TDATR adopts a ``perceive-then-fuse'' strategy. The model first performs table detail-aware learning to jointly perceive table structure and content through multiple structure understanding and content recognition tasks designed under a language modeling paradigm. These tasks can naturally leverage document data from diverse scenarios to enhance model robustness. The model then integrates implicit table details to generate structured HTML outputs, enabling more efficient TR modeling when trained with limited data. Furthermore, we design a structure-guided cell localization module integrated into the end-to-end TR framework, which efficiently locates cell and strengthens vision-language alignment. It enhances the interpretability and accuracy of TR. We achieve state-of-the-art or highly competitive performance on seven benchmarks without dataset-specific fine-tuning.

cs.CV↗

arXiv2026

VSD-MOT: End-to-End Multi-Object Tracking in Low-Quality Video Scenes Guided by Visual Semantic Distillation

Existing multi-object tracking algorithms typically fail to adequately address the issues in low-quality videos, resulting in a significant decline in tracking performance when image quality deteriorates in real-world scenarios. This performance degradation is primarily due to the algorithms' inability to effectively tackle the problems caused by information loss in low-quality images. To address the challenges of low-quality video scenarios, inspired by vision-language models, we propose a multi-object tracking framework guided by visual semantic distillation (VSD-MOT). Specifically, we introduce the CLIP Image Encoder to extract global visual semantic information from images to compensate for the loss of information in low-quality images. However, direct integration can substantially impact the efficiency of the multi-object tracking algorithm. Therefore, this paper proposes to extract visual semantic information from images through knowledge distillation. This method adopts a teacher-student learning framework, with the CLIP Image Encoder serving as the teacher model. To enable the student model to acquire the capability of extracting visual semantic information suitable for multi-object tracking tasks from the teacher model, we have designed the Dual-Constraint Semantic Distillation method (DCSD). Furthermore, to address the dynamic variation of frame quality in low-quality videos, we propose the Dynamic Semantic Weight Regulation (DSWR) module, which adaptively allocates fusion weights based on real-time frame quality assessment. Extensive experiments demonstrate the effectiveness and superiority of the proposed method in low-quality video scenarios in the real world. Meanwhile, our method can maintain good performance in conventional scenarios.

cs.CV↗

arXiv2026

EARTalking: End-to-end GPT-style Autoregressive Talking Head Synthesis with Frame-wise Control

Audio-driven talking head generation aims to create vivid and realistic videos from a static portrait and speech. Existing AR-based methods rely on intermediate facial representations, which limit their expressiveness and realism. Meanwhile, diffusion-based methods generate clip-by-clip, lacking fine-grained control and causing inherent latency due to overall denoising across the window. To address these limitations, we propose EARTalking, a novel end-to-end, GPT-style autoregressive model for interactive audio-driven talking head generation. Our method introduces a novel frame-by-frame, in-context, audio-driven streaming generation paradigm. For inherently supporting variable-length video generation with identity consistency, we propose the Sink Frame Window Attention (SFA) mechanism. Furthermore, to avoid the complex, separate networks that prior works required for diverse control signals, we propose a streaming Frame Condition In-Context (FCIC) scheme. This scheme efficiently injects diverse control signals in a streaming, in-context manner, enabling interactive control at every frame and at arbitrary moments. Experiments demonstrate that EARTalking outperforms existing autoregressive methods and achieves performance comparable to diffusion-based methods. Our work demonstrates the feasibility of in-context streaming autoregressive control, unlocking a scalable direction for flexible, efficient generation. The code will be released for reproducibility.

cs.CV↗

arXiv2026

A Stable, High-Order Time-Stepping Scheme for the Drift-Diffusion Model in Modern Solar Cell Simulation

This paper presents a one-dimensional transient drift--diffusion simulator for advanced solar cells, integrating a structure-preserving finite-volume spatial discretization with Scharfetter--Gummel--type fluxes and a high-order, L-stable implicit Runge--Kutta (Radau IIA) temporal integrator. The scheme ensures local charge conservation, handles sharp material interfaces, and achieves second-order spatial and fifth-order temporal convergence. Its accuracy is verified against the classical depletion approximation in $p$--$n$ junction and validated through excellent agreement with the established simulator for an organic photovoltaic device. The framework's extensibility is demonstrated by incorporating exciton kinetics in organic solar cells, capturing multi-timescale dynamics, and by modeling mobile ions in perovskite solar cells, reproducing characteristic $\tmem{J}$--$\tmem{V}$ hysteresis without empirical parameters. This work provides a robust, high-order numerical foundation for simulating coupled charge, exciton, and ion transport in next-generation photovoltaic devices.

physics.app-ph↗

arXiv2026

Multi-Domain Audio Question Answering Benchmark Toward Acoustic Content Reasoning

We present Task 5 of the DCASE 2025 Challenge: an Audio Question Answering (AQA) benchmark spanning multiple domains of sound understanding. This task defines three QA subsets (Bioacoustics, Temporal Soundscapes, and Complex QA) to test audio-language models on interactive question-answering over diverse acoustic scenes. We describe the dataset composition (from marine mammal calls to soundscapes and complex real-world clips), the evaluation protocol (top-1 accuracy with answer-shuffling robustness), and baseline systems (Qwen2-Audio-7B, AudioFlamingo 2, Gemini-2-Flash). Preliminary results on the development set are compared, showing strong variation across models and subsets. This challenge aims to advance the audio understanding and reasoning capabilities of audio-language models toward human-level acuity, which are crucial for enabling AI agents to perceive and interact about the world effectively.

cs.SD↗

arXiv2026

The USTC-NERCSLIP Systems for the CHiME-9 MCoRec Challenge

This report details our submission to the CHiME-9 MCoRec Challenge on recognizing and clustering multiple concurrent natural conversations within indoor social settings. Unlike conventional meetings centered on a single shared topic, this scenario contains multiple parallel dialogues--up to eight speakers across up to four simultaneous conversations--with a speech overlap rate exceeding 90%. To tackle this, we propose a multimodal cascaded system that leverages per-speaker visual streams extracted from synchronized 360 degree video together with single-channel audio. Our system improves three components of the pipeline by leveraging enhanced audio-visual pretrained models: Active Speaker Detection (ASD), Audio-Visual Target Speech Extraction (AVTSE), and Audio-Visual Speech Recognition (AVSR). The AVSR module further incorporates Whisper and LLM techniques to boost transcription accuracy. Our best single cascaded system achieves a Speaker Word Error Rate (WER) of 32.44% on the development set. By further applying ROVER to fuse outputs from diverse front-end and back-end variants, we reduce Speaker WER to 31.40%. Notably, our LLM-based zero-shot conversational clustering achieves a speaker clustering F1 score of 1.0, yielding a final Joint ASR-Clustering Error Rate (JACER) of 15.70%.

eess.AS↗

arXiv2026

THOR: Tool-Integrated Hierarchical Optimization via RL for Mathematical Reasoning

Large Language Models (LLMs) have made remarkable progress in mathematical reasoning, but still continue to struggle with high-precision tasks like numerical computation and formal symbolic manipulation. Integrating external tools has emerged as a promising approach to bridge this gap. Despite recent advances, existing methods struggle with three key challenges: constructing tool-integrated reasoning data, performing fine-grained optimization, and enhancing inference. To overcome these limitations, we propose THOR (Tool-Integrated Hierarchical Optimization via RL). First, we introduce TIRGen, a multi-agent based pipeline for constructing high-quality datasets of tool-integrated reasoning paths, aligning with the policy and generalizing well across diverse models. Second, to perform fine-grained hierarchical optimization, we introduce an RL strategy that jointly optimizes for both episode-level problem solving and step-level code generation. This is motivated by our key insight that the success of an intermediate tool call is a strong predictor of the final answer's correctness. Finally, THOR incorporates a self-correction mechanism that leverages immediate tool feedback to dynamically revise erroneous reasoning paths during inference. Our approach demonstrates strong generalization across diverse models, performing effectively in both reasoning and non-reasoning models. It further achieves state-of-the-art performance for models of a similar scale on multiple mathematical benchmarks, while also delivering consistent improvements on code benchmarks. Our code will be publicly available at https://github.com/JingMog/THOR.

cs.AI↗

arXiv2026

Split Federated Learning for Low-Altitude Wireless Networks: Joint Sensing, Communication, Computation, and Control Co-design

Unmanned aerial vehicles (UAVs) with integrated sensing, communication, computation and control (ISC3) capabilities have become key enablers of next-generation wireless networks. Federated edge learning (FEL) leverages UAVs as mobile learning agents to collect data, perform local model updates, and contribute to global model aggregation. However, existing UAV-assisted FEL systems face critical challenges, including excessive computational demands, privacy risks, and inefficient communication, primarily due to the requirement for full-model training on resource-constrained UAVs. To deal with aforementioned challenges, we propose Split Federated Learning for UAV-Enabled ISC3 (SFLSC3), a novel framework that integrates split federated learning (SFL) into UAV-assisted FEL. SFLSC3 optimally partitions model training between UAVs and edge servers, significantly reducing UAVs' computational burden while preserving data privacy. We conduct a theoretical analysis of UAV deployment, split point selection, data sensing volume, and client-side aggregation frequency, deriving closed-form upper bounds for the convergence gap. Based on these insights, we conceive a joint optimization problem to minimize the delay required to achieve a target model accuracy. Given the non-convex nature of the problem, we develop a low-complexity algorithm to efficiently determine UAV deployment, split point selection, and communication frequency. Extensive simulations on a target motion recognition task validate the effectiveness of SFLSC3, demonstrating superior convergence and delay performance compared to baseline methods.

cs.DC↗

arXiv2026

REST: Diffusion-based Real-time End-to-end Streaming Talking Head Generation via ID-Context Caching and Asynchronous Streaming Distillation

Diffusion models have significantly advanced the field of talking head generation (THG). However, slow inference speeds and prevalent non-autoregressive paradigms severely constrain the application of diffusion-based THG models. In this study, we propose REST, a pioneering diffusion-based, real-time, end-to-end streaming audio-driven talking head generation framework. To support real-time end-to-end generation, a compact video latent space is first learned through a spatiotemporal variational autoencoder with a high compression ratio. Additionally, to enable semi-autoregressive streaming within the compact video latent space, we introduce an ID-Context Cache mechanism, which integrates ID-Sink and Context-Cache principles into key-value caching for maintaining identity consistency and temporal coherence during long-term streaming generation. Furthermore, an Asynchronous Streaming Distillation (ASD) strategy is proposed to mitigate error accumulation and enhance temporal consistency in streaming generation, leveraging a non-streaming teacher with an asynchronous noise schedule to supervise the streaming student. REST bridges the gap between autoregressive and diffusion-based approaches, achieving a breakthrough in efficiency for applications requiring real-time THG. Experimental results demonstrate that REST outperforms state-of-the-art methods in both generation speed and overall performance.

cs.CV↗

arXiv2026

Step Potential Advantage Estimation: Harnessing Intermediate Confidence and Correctness for Efficient Mathematical Reasoning

Reinforcement Learning with Verifiable Rewards (RLVR) elicits long chain-of-thought reasoning in large language models (LLMs), but outcome-based rewards lead to coarse-grained advantage estimation. While existing approaches improve RLVR via token-level entropy or sequence-level length control, they lack a semantically grounded, step-level measure of reasoning progress. As a result, LLMs fail to distinguish necessary deduction from redundant verification: they may continue checking after reaching a correct solution and, in extreme cases, overturn a correct trajectory into an incorrect final answer. To remedy the lack of process supervision, we introduce a training-free probing mechanism that extracts intermediate confidence and correctness and combines them into a Step Potential signal that explicitly estimates the reasoning state at each step. Building on this signal, we propose Step Potential Advantage Estimation (SPAE), a fine-grained credit assignment method that amplifies potential gains, penalizes potential drops, and applies penalty after potential saturates to encourage timely termination. Experiments across multiple benchmarks show SPAE consistently improves accuracy while substantially reducing response length, outperforming strong RL baselines and recent efficient reasoning and token-level advantage estimation methods. The code is available at https://github.com/cii030/SPAE-RL.

cs.CL↗

arXiv2026

Rethinking Secure Semantic Communications in the Age of Generative and Agentic AI: Threats and Opportunities

Semantic communication (SemCom) improves communication efficiency by transmitting task-relevant information instead of raw bits and is expected to be a key technology for 6G networks. Recent advances in generative AI (GenAI) further enhance SemCom by enabling robust semantic encoding and decoding under limited channel conditions. However, these efficiency gains also introduce new security and privacy vulnerabilities. Due to the broadcast nature of wireless channels, eavesdroppers can also use powerful GenAI-based semantic decoders to recover private information from intercepted signals. Moreover, rapid advances in agentic AI enable eavesdroppers to perform long-term and adaptive inference through the integration of memory, external knowledge, and reasoning capabilities. This allows eavesdroppers to further infer user private behavior and intent beyond the transmitted content. Motivated by these emerging challenges, this paper comprehensively rethinks the security and privacy of SemCom systems in the age of generative and agentic AI. We first present a systematic taxonomy of eavesdropping threat models in SemCom systems. Then, we provide insights into how GenAI and agentic AI can enhance eavesdropping threats. Meanwhile, we also highlight potential opportunities for leveraging GenAI and agentic AI to design privacy-preserving SemCom systems.

eess.SP↗

arXiv2025

KH-FUNSD: A Hierarchical and Fine-Grained Layout Analysis Dataset for Low-Resource Khmer Business Document

Automated document layout analysis remains a major challenge for low-resource, non-Latin scripts. Khmer is a language spoken daily by over 17 million people in Cambodia, receiving little attention in the development of document AI tools. The lack of dedicated resources is particularly acute for business documents, which are critical for both public administration and private enterprise. To address this gap, we present \textbf{KH-FUNSD}, the first publicly available, hierarchically annotated dataset for Khmer form document understanding, including receipts, invoices, and quotations. Our annotation framework features a three-level design: (1) region detection that divides each document into core zones such as header, form field, and footer; (2) FUNSD-style annotation that distinguishes questions, answers, headers, and other key entities, together with their relationships; and (3) fine-grained classification that assigns specific semantic roles, such as field labels, values, headers, footers, and symbols. This multi-level approach supports both comprehensive layout analysis and precise information extraction. We benchmark several leading models, providing the first set of baseline results for Khmer business documents, and discuss the distinct challenges posed by non-Latin, low-resource scripts. The KH-FUNSD dataset and documentation will be available at URL.

cs.CL↗

arXiv2025

Binary-Gaussian: Compact and Progressive Representation for 3D Gaussian Segmentation

3D Gaussian Splatting (3D-GS) has emerged as an efficient 3D representation and a promising foundation for semantic tasks like segmentation. However, existing 3D-GS-based segmentation methods typically rely on high-dimensional category features, which introduce substantial memory overhead. Moreover, fine-grained segmentation remains challenging due to label space congestion and the lack of stable multi-granularity control mechanisms. To address these limitations, we propose a coarse-to-fine binary encoding scheme for per-Gaussian category representation, which compresses each feature into a single integer via the binary-to-decimal mapping, drastically reducing memory usage. We further design a progressive training strategy that decomposes panoptic segmentation into a series of independent sub-tasks, reducing inter-class conflicts and thereby enhancing fine-grained segmentation capability. Additionally, we fine-tune opacity during segmentation training to address the incompatibility between photometric rendering and semantic segmentation, which often leads to foreground-background confusion. Extensive experiments on multiple benchmarks demonstrate that our method achieves state-of-the-art segmentation performance while significantly reducing memory consumption and accelerating inference.

cs.CV↗

arXiv2025

Graph-Aware Temporal Encoder Based Service Migration and Resource Allocation in Satellite Networks

The rapid expansion of latency-sensitive applications has sparked renewed interest in deploying edge computing capabilities aboard satellite constellations, aiming to achieve truly global and seamless service coverage. On one hand, it is essential to allocate the limited onboard computational and communication resources efficiently to serve geographically distributed users. On the other hand, the dynamic nature of satellite orbits necessitates effective service migration strategies to maintain service continuity and quality as the coverage areas of satellites evolve. We formulate this problem as a spatio-temporal Markov decision process, where satellites, ground users, and flight users are modeled as nodes in a time-varying graph. The node features incorporate queuing dynamics to characterize packet loss probabilities. To solve this problem, we propose a Graph-Aware Temporal Encoder (GATE) that jointly models spatial correlations and temporal dynamics. GATE uses a two-layer graph convolutional network to extract inter-satellite and user dependencies and a temporal convolutional network to capture their short-term evolution, producing unified spatio-temporal representations. The resulting spatial-temporal representations are passed into a Hybrid Proximal Policy Optimization (HPPO) framework. This framework features a multi-head actor that outputs both discrete service migration decisions and continuous resource allocation ratios, along with a critic for value estimation. We conduct extensive simulations involving both persistent and intermittent users distributed across real-world population centers.

cs.NI↗

arXiv2025

READ: Real-time and Efficient Asynchronous Diffusion for Audio-driven Talking Head Generation

The introduction of diffusion models has brought significant advances to the field of audio-driven talking head generation. However, the extremely slow inference speed severely limits the practical implementation of diffusion-based talking head generation models. In this study, we propose READ, a real-time diffusion-transformer-based talking head generation framework. Our approach first learns a spatiotemporal highly compressed video latent space via a temporal VAE, significantly reducing the token count to accelerate generation. To achieve better audio-visual alignment within this compressed latent space, a pre-trained Speech Autoencoder (SpeechAE) is proposed to generate temporally compressed speech latent codes corresponding to the video latent space. These latent representations are then modeled by a carefully designed Audio-to-Video Diffusion Transformer (A2V-DiT) backbone for efficient talking head synthesis. Furthermore, to ensure temporal consistency and accelerated inference in extended generation, we propose a novel asynchronous noise scheduler (ANS) for both the training and inference processes of our framework. The ANS leverages asynchronous add-noise and asynchronous motion-guided generation in the latent space, ensuring consistency in generated video clips. Experimental results demonstrate that READ outperforms state-of-the-art methods by generating competitive talking head videos with significantly reduced runtime, achieving an optimal balance between quality and speed while maintaining robust metric stability in long-time generation.

cs.GR↗

arXiv2025

From Structure to Detail: Hierarchical Distillation for Efficient Diffusion Model

The inference latency of diffusion models remains a critical barrier to their real-time application. While trajectory-based and distribution-based step distillation methods offer solutions, they present a fundamental trade-off. Trajectory-based methods preserve global structure but act as a "lossy compressor", sacrificing high-frequency details. Conversely, distribution-based methods can achieve higher fidelity but often suffer from mode collapse and unstable training. This paper recasts them from independent paradigms into synergistic components within our novel Hierarchical Distillation (HD) framework. We leverage trajectory distillation not as a final generator, but to establish a structural ``sketch", providing a near-optimal initialization for the subsequent distribution-based refinement stage. This strategy yields an ideal initial distribution that enhances the ceiling of overall performance. To further improve quality, we introduce and refine the adversarial training process. We find standard discriminator structures are ineffective at refining an already high-quality generator. To overcome this, we introduce the Adaptive Weighted Discriminator (AWD), tailored for the HD pipeline. By dynamically allocating token weights, AWD focuses on local imperfections, enabling efficient detail refinement. Our approach demonstrates state-of-the-art performance across diverse tasks. On ImageNet $256\times256$, our single-step model achieves an FID of 2.26, rivaling its 250-step teacher. It also achieves promising results on the high-resolution text-to-image MJHQ benchmark, proving its generalizability. Our method establishes a robust new paradigm for high-fidelity, single-step diffusion models.

cs.CV↗

arXiv2025

RIS-based Communication Enhancement and Location Privacy Protection in UAV Networks

With the explosive advancement of unmanned aerial vehicles (UAVs), the security of efficient UAV networks has become increasingly critical. Owing to the open nature of its communication environment, illegitimate malicious UAVs (MUs) can infer the position of the source UAV (SU) by analyzing received signals, thus compromising the SU location privacy. To protect the SU location privacy while ensuring efficient communication with legitimate receiving UAVs (RUs), we propose an Active Reconfigurable Intelligent Surface (ARIS)-assisted covert communication scheme based on virtual partitioning and artificial noise (AN). Specifically, we design a novel ARIS architecture integrated with an AN module. This architecture dynamically partitions its reflecting elements into multiple sub-regions: one subset is optimized to enhance the communication rate between the SU and RUs, while the other subset generates AN to interfere with the localization of the SU by MUs. We first derive the Cramér-Rao Lower Bound (CRLB) for the localization with received signal strength (RSS), based on which, we establish a joint optimization framework for communication enhancement and localization interference. Subsequently, we derive and validate the optimal ARIS partitioning and power allocation under average channel conditions. Finally, tailored optimization methods are proposed for the reflection precoding and AN design of the two partitions. Simulation results validate that, compared to baseline schemes, the proposed scheme significantly increases the localization error of the SU by MUs while maintaining efficient communication between the SU and RUs, thereby effectively protecting the SU location privacy.

eess.SY↗