arXiv subjects

Zekun Li

Publications and source records attributed to Zekun Li.

At least 55 records · Page 3Linked to original sources

arXiv2026

EgoReAct: Egocentric Video-Driven 3D Human Reaction Generation

Humans exhibit adaptive, context-sensitive responses to egocentric visual input. However, faithfully modeling such reactions from egocentric video remains challenging due to the dual requirements of strictly causal generation and precise 3D spatial alignment. To tackle this problem, we first construct the Human Reaction Dataset (HRD) to address data scarcity and misalignment by building a spatially aligned egocentric video-reaction dataset, as existing datasets (e.g., ViMo) suffer from significant spatial inconsistency between the egocentric video and reaction motion, e.g., dynamically moving motions are always paired with fixed-camera videos. Leveraging HRD, we present EgoReAct, the first autoregressive framework that generates 3D-aligned human reaction motions from egocentric video streams in real-time. We first compress the reaction motion into a compact yet expressive latent space via a Vector Quantised-Variational AutoEncoder and then train a Generative Pre-trained Transformer for reaction generation from the visual input. EgoReAct incorporates 3D dynamic features, i.e., metric depth, and head dynamics during the generation, which effectively enhance spatial grounding. Extensive experiments demonstrate that EgoReAct achieves remarkably higher realism, spatial consistency, and generation efficiency compared with prior methods, while maintaining strict causality during generation. We will release code, models, and data upon acceptance.

cs.CV↗

arXiv2025

T^2Agent A Tool-augmented Multimodal Misinformation Detection Agent with Monte Carlo Tree Search

Real-world multimodal misinformation often arises from mixed forgery sources, requiring dynamic reasoning and adaptive verification. However, existing methods mainly rely on static pipelines and limited tool usage, limiting their ability to handle such complexity and diversity. To address this challenge, we propose \method, a novel misinformation detection agent that incorporates an extensible toolkit with Monte Carlo Tree Search (MCTS). The toolkit consists of modular tools such as web search, forgery detection, and consistency analysis. Each tool is described using standardized templates, enabling seamless integration and future expansion. To avoid inefficiency from using all tools simultaneously, a greedy search-based selector is proposed to identify a task-relevant subset. This subset then serves as the action space for MCTS to dynamically collect evidence and perform multi-source verification. To better align MCTS with the multi-source nature of misinformation detection, \method~ extends traditional MCTS with multi-source verification, which decomposes the task into coordinated subtasks targeting different forgery sources. A dual reward mechanism containing a reasoning trajectory score and a confidence score is further proposed to encourage a balance between exploration across mixed forgery sources and exploitation for more reliable evidence. We conduct ablation studies to confirm the effectiveness of the tree search mechanism and tool usage. Extensive experiments further show that \method~ consistently outperforms existing baselines on challenging mixed-source multimodal misinformation benchmarks, demonstrating its strong potential as a training-free detector.

cs.CL↗

arXiv2025

Video-SafetyBench: A Benchmark for Safety Evaluation of Video LVLMs

The increasing deployment of Large Vision-Language Models (LVLMs) raises safety concerns under potential malicious inputs. However, existing multimodal safety evaluations primarily focus on model vulnerabilities exposed by static image inputs, ignoring the temporal dynamics of video that may induce distinct safety risks. To bridge this gap, we introduce Video-SafetyBench, the first comprehensive benchmark designed to evaluate the safety of LVLMs under video-text attacks. It comprises 2,264 video-text pairs spanning 48 fine-grained unsafe categories, each pairing a synthesized video with either a harmful query, which contains explicit malice, or a benign query, which appears harmless but triggers harmful behavior when interpreted alongside the video. To generate semantically accurate videos for safety evaluation, we design a controllable pipeline that decomposes video semantics into subject images (what is shown) and motion text (how it moves), which jointly guide the synthesis of query-relevant videos. To effectively evaluate uncertain or borderline harmful outputs, we propose RJScore, a novel LLM-based metric that incorporates the confidence of judge models and human-aligned decision threshold calibration. Extensive experiments show that benign-query video composition achieves average attack success rates of 67.2%, revealing consistent vulnerabilities to video-induced attacks. We believe Video-SafetyBench will catalyze future research into video-based safety evaluation and defense strategies.

cs.CV↗

arXiv2025

SpineBench: Benchmarking Multimodal LLMs for Spinal Pathology Analysis

With the increasing integration of Multimodal Large Language Models (MLLMs) into the medical field, comprehensive evaluation of their performance in various medical domains becomes critical. However, existing benchmarks primarily assess general medical tasks, inadequately capturing performance in nuanced areas like the spine, which relies heavily on visual input. To address this, we introduce SpineBench, a comprehensive Visual Question Answering (VQA) benchmark designed for fine-grained analysis and evaluation of MLLMs in the spinal domain. SpineBench comprises 64,878 QA pairs from 40,263 spine images, covering 11 spinal diseases through two critical clinical tasks: spinal disease diagnosis and spinal lesion localization, both in multiple-choice format. SpineBench is built by integrating and standardizing image-label pairs from open-source spinal disease datasets, and samples challenging hard negative options for each VQA pair based on visual similarity (similar but not the same disease), simulating real-world challenging scenarios. We evaluate 12 leading MLLMs on SpineBench. The results reveal that these models exhibit poor performance in spinal tasks, highlighting limitations of current MLLM in the spine domain and guiding future improvements in spinal medicine applications. SpineBench is publicly available at https://zhangchenghanyu.github.io/SpineBench.github.io/.

cs.CV↗

arXiv2025

Beyond the Strongest LLM: Multi-Turn Multi-Agent Orchestration vs. Single LLMs on Benchmarks

We study multi-turn multi-agent orchestration, where multiple large language model (LLM) agents interact over multiple turns by iteratively proposing answers or casting votes until reaching consensus. Using four LLMs (Gemini 2.5 Pro, GPT-5, Grok 4, and Claude Sonnet 4) on GPQA-Diamond, IFEval, and MuSR, we conduct two experiments: (i) benchmarking orchestration against single-LLM baselines; and (ii) ablations on GPQA-Diamond that vary whether agents see who authored answers and whether they can observe ongoing votes. Orchestration matches or exceeds the strongest single model and consistently outperforms the others. Analysis of best-achievable orchestration performance shows potential for further gains. The ablations show that revealing authorship increases self-voting and ties, and that showing ongoing votes amplifies herding, which speeds convergence but can sometimes yield premature consensus.

cs.AI↗

arXiv2025

Correspondence as Video: Test-Time Adaption on SAM2 for Reference Segmentation in the Wild

Large vision models like the Segment Anything Model (SAM) exhibit significant limitations when applied to downstream tasks in the wild. Consequently, reference segmentation, which leverages reference images and their corresponding masks to impart novel knowledge to the model, emerges as a promising new direction for adapting vision models. However, existing reference segmentation approaches predominantly rely on meta-learning, which still necessitates an extensive meta-training process and brings massive data and computational cost. In this study, we propose a novel approach by representing the inherent correspondence between reference-target image pairs as a pseudo video. This perspective allows the latest version of SAM, known as SAM2, which is equipped with interactive video object segmentation (iVOS) capabilities, to be adapted to downstream tasks in a lightweight manner. We term this approach Correspondence As Video for SAM (CAV-SAM). CAV-SAM comprises two key modules: the Diffusion-Based Semantic Transition (DBST) module employs a diffusion model to construct a semantic transformation sequence, while the Test-Time Geometric Alignment (TTGA) module aligns the geometric changes within this sequence through test-time fine-tuning. We evaluated CAVSAM on widely-used datasets, achieving segmentation performance improvements exceeding 5% over SOTA methods. Implementation is provided in the supplementary materials.

cs.CV↗

arXiv2025

Rethinking the Role of Prompting Strategies in LLM Test-Time Scaling: A Perspective of Probability Theory

Recently, scaling test-time compute on Large Language Models (LLM) has garnered wide attention. However, there has been limited investigation of how various reasoning prompting strategies perform as scaling. In this paper, we focus on a standard and realistic scaling setting: majority voting. We systematically conduct experiments on 6 LLMs $\times$ 8 prompting strategies $\times$ 6 benchmarks. Experiment results consistently show that as the sampling time and computational overhead increase, complicated prompting strategies with superior initial performance gradually fall behind simple Chain-of-Thought. We analyze this phenomenon and provide theoretical proofs. Additionally, we propose a probabilistic method to efficiently predict scaling performance and identify the best prompting strategy under large sampling times, eliminating the need for resource-intensive inference processes in practical applications. Furthermore, we introduce two ways derived from our theoretical analysis to significantly improve the scaling performance. We hope that our research can promote to re-examine the role of complicated prompting, unleash the potential of simple prompting strategies, and provide new insights for enhancing test-time scaling performance. Code is available at https://github.com/MraDonkey/rethinking_prompting.

cs.AI↗

arXiv2025

Development and in silico imaging trial evaluation of a deep-learning-based transmission-less attenuation compensation method for DaT SPECT

Quantitative measures of dopamine transporter (DaT) uptake in caudate, putamen, and globus pallidus derived from DaT-single-photon emission computed tomography (SPECT) images are being investigated as biomarkers to diagnose, assess disease status, and track the progression of Parkinsonism. Reliable quantification from DaT-SPECT images requires performing attenuation compensation (AC), typically with a separate X-ray CT scan. Such CT-based AC (CTAC) has multiple challenges, a key one being the non-availability of X-ray CT component on many clinical SPECT systems. Even when a CT is available, the additional CT scan leads to increased radiation dose, costs, and complexity, potential quantification errors due to SPECT-CT misalignment, and higher training and regulatory requirements. To overcome the challenges with the requirement of a CT scan for AC in DaT SPECT, we propose a deep learning (DL)-based transmission-less AC method for DaT-SPECT (DaT-CTLESS). An in silico imaging trial, titled ISIT-DaT, was designed to evaluate the performance of DaT-CTLESS on the regional uptake quantification task. We observed that DaT-CTLESS yielded a significantly higher correlation with CTAC than that between UAC and CTAC on the regional DaT uptake quantification task. Further, DaT-CLTESS had an excellent agreement with CTAC on this task, significantly outperformed UAC in distinguishing patients with normal versus reduced putamen SBR, yielded good generalizability across two scanners, was generally insensitive to intra-regional uptake heterogeneity, demonstrated good repeatability, exhibited robust performance even as the size of the training data was reduced, and generally outperformed the other considered DL methods on the task of quantifying regional uptake across different training dataset sizes. These results provide a strong motivation for further clinical evaluation of DaT-CTLESS.

physics.med-ph↗

arXiv2025

DIGMAPPER: A Modular System for Automated Geologic Map Digitization

Historical geologic maps contain rich geospatial information, such as rock units, faults, folds, and bedding planes, that is critical for assessing mineral resources essential to renewable energy, electric vehicles, and national security. However, digitizing maps remains a labor-intensive and time-consuming task. We present DIGMAPPER, a modular, scalable system developed in collaboration with the United States Geological Survey (USGS) to automate the digitization of geologic maps. DIGMAPPER features a fully dockerized, workflow-orchestrated architecture that integrates state-of-the-art deep learning models for map layout analysis, feature extraction, and georeferencing. To overcome challenges such as limited training data and complex visual content, our system employs innovative techniques, including in-context learning with large language models, synthetic data generation, and transformer-based models. Evaluations on over 100 annotated maps from the DARPA-USGS dataset demonstrate high accuracy across polygon, line, and point feature extraction, and reliable georeferencing performance. Deployed at USGS, DIGMAPPER significantly accelerates the creation of analysis-ready geospatial datasets, supporting national-scale critical mineral assessments and broader geoscientific applications.

cs.CV↗

arXiv2025

SOPBench: Evaluating Language Agents at Following Standard Operating Procedures and Constraints

As language agents increasingly automate critical tasks, their ability to follow domain-specific standard operating procedures (SOPs), policies, and constraints when taking actions and making tool calls becomes essential yet remains underexplored. To address this gap, we develop an automated evaluation pipeline SOPBench with: (1) executable environments containing 167 tools/functions across seven customer service domains with service-specific SOPs and rule-based verifiers, (2) an automated test generation framework producing over 900 verified test cases, and (3) an automated evaluation framework to rigorously assess agent adherence from multiple dimensions. Our approach transforms each service-specific SOP code program into a directed graph of executable functions and requires agents to call these functions based on natural language SOP descriptions. The original code serves as oracle rule-based verifiers to assess compliance, reducing reliance on manual annotations and LLM-based evaluations. We evaluate 18 leading models, and results show the task is challenging even for top-tier models (like GPT-4o, Claude-3.7-Sonnet), with variances across domains. Reasoning models like o4-mini-high show superiority while other powerful models perform less effectively (pass rates of 30%-50%), and small models (7B, 8B) perform significantly worse. Additionally, language agents can be easily jailbroken to overlook SOPs and constraints. Code, data, and over 24k agent trajectories are released at https://github.com/Leezekun/SOPBench.

cs.CL↗

arXiv2025

The Latency Price of Threshold Cryptosystem in Blockchains

Threshold cryptography is essential for many blockchain protocols. For example, many protocols rely on threshold common coin to implement asynchronous consensus, leader elections, and provide support for randomized applications. Similarly, threshold decryption and threshold time-lock puzzles are often necessary for privacy. In this paper, we study the interplay between threshold cryptography and a class of blockchains that use Byzantine-fault tolerant (BFT) consensus protocols with a focus on latency. More specifically, we focus on blockchain-native threshold cryptosystem, where the blockchain validators seek to run a threshold cryptographic protocol once for every block with the block contents as an input to the threshold cryptographic protocol. All existing approaches for blockchain-native threshold cryptosystems introduce a latency overhead of at least one message delay for running the threshold cryptographic protocol. In this paper, we first propose a mechanism to eliminate this overhead for blockchain-native threshold cryptosystems with tight thresholds, i.e., in threshold cryptographic protocols where the secrecy and reconstruction thresholds are the same. However, many real-world proof-of-stake-based blockchain-native threshold cryptosystems rely on ramp thresholds, where reconstruction thresholds are strictly greater than secrecy thresholds. For these blockchains, we formally demonstrate that the additional delay is unavoidable. We then introduce a mechanism to minimize this delay in the optimistic case. We implement our optimistic protocol for the proof-of-stake distributed randomness scheme on the Aptos blockchain. Our measurements from the Aptos mainnet show that the optimistic approach reduces latency overhead by 71%.

cs.CR↗

arXiv2025

MacRAG: Compress, Slice, and Scale-up for Multi-Scale Adaptive Context RAG

Long-context large language models (LC LLMs) combined with retrieval-augmented generation (RAG) hold strong potential for complex multi-hop and large-document tasks. However, existing RAG systems often suffer from imprecise retrieval, incomplete context coverage under constrained windows, and fragmented information from suboptimal context construction. We introduce Multi-scale Adaptive Context RAG (MacRAG), a hierarchical RAG framework that compresses and partitions documents into coarse-to-fine granularities, then adaptively merges relevant contexts through real-time chunk- and document-level expansions. By initiating with finest-level retrieval and progressively incorporating broader, higher-level context, MacRAG constructs effective query-specific long contexts, optimizing both precision and coverage. Evaluations on challenging LongBench expansions of HotpotQA, 2WikiMultihopQA, and Musique confirm MacRAG consistently surpasses baseline RAG pipelines in single- and multi-step generation using Llama-3.1-8B, Gemini-1.5-pro, and GPT-4o. Our results establish MacRAG as an efficient, scalable solution for real-world long-context, multi-hop reasoning. Our code is available at https://github.com/Leezekun/MacRAG.

cs.CL↗

arXiv2025

Replace Anyone in Videos

The field of controllable human-centric video generation has witnessed remarkable progress, particularly with the advent of diffusion models. However, achieving precise and localized control over human motion in videos, such as replacing or inserting individuals while preserving desired motion patterns, still remains a formidable challenge. In this work, we present the ReplaceAnyone framework, which focuses on localized human replacement and insertion featuring intricate backgrounds. Specifically, we formulate this task as an image-conditioned video inpainting paradigm with pose guidance, utilizing a unified end-to-end video diffusion architecture that facilitates image-conditioned video inpainting within masked regions. To prevent shape leakage and enable granular local control, we introduce diverse mask forms involving both regular and irregular shapes. Furthermore, we implement an enriched visual guidance mechanism to enhance appearance alignment, a hybrid inpainting encoder to further preserve the detailed background information in the masked video, and a two-phase optimization methodology to simplify the training difficulty. ReplaceAnyone enables seamless replacement or insertion of characters while maintaining the desired pose motion and reference appearance within a single framework. Extensive experimental results demonstrate the effectiveness of our method in generating realistic and coherent video content. The proposed ReplaceAnyone can be seamlessly applied not only to traditional 3D-UNet base models but also to DiT-based video models such as Wan2.1. The code will be available at https://github.com/ali-vilab/UniAnimate-DiT.

cs.CV↗

arXiv2025

Raptr: Prefix Consensus for Robust High-Performance BFT

In this paper, we present Raptr--a Byzantine fault-tolerant state machine replication (BFT SMR) protocol that combines strong robustness with high throughput, while attaining near-optimal theoretical latency. Raptr delivers exceptionally low latency and high throughput under favorable conditions, and it degrades gracefully in the presence of Byzantine faults and network attacks. Existing high-throughput BFT SMR protocols typically take either pessimistic or optimistic approaches to data dissemination: the former suffers from suboptimal latency in favorable conditions, while the latter deteriorates sharply under minimal attacks or network instability. Raptr bridges this gap, combining the strengths of both approaches through a novel Prefix Consensus mechanism. We implement Raptr and evaluate it against several state-of-the-art protocols in a geo-distributed environment with 100 replicas. Raptr achieves 260,000 transactions per second (TPS) with sub-second latency under favorable conditions, sustaining 610ms at 10,000 TPS and 755ms at 250,000 TPS. It remains robust under network glitches, showing minimal performance degradation even with a 1% message drop rate.

cs.DC↗

arXiv2025

Balancing Multi-Target Semi-Supervised Medical Image Segmentation with Collaborative Generalist and Specialists

Despite the promising performance achieved by current semi-supervised models in segmenting individual medical targets, many of these models suffer a notable decrease in performance when tasked with the simultaneous segmentation of multiple targets. A vital factor could be attributed to the imbalanced scales among different targets: during simultaneously segmenting multiple targets, large targets dominate the loss, leading to small targets being misclassified as larger ones. To this end, we propose a novel method, which consists of a Collaborative Generalist and several Specialists, termed CGS. It is centered around the idea of employing a specialist for each target class, thus avoiding the dominance of larger targets. The generalist performs conventional multi-target segmentation, while each specialist is dedicated to distinguishing a specific target class from the remaining target classes and the background. Based on a theoretical insight, we demonstrate that CGS can achieve a more balanced training. Moreover, we develop cross-consistency losses to foster collaborative learning between the generalist and the specialists. Lastly, regarding their intrinsic relation that the target class of any specialized head should belong to the remaining classes of the other heads, we introduce an inter-head error detection module to further enhance the quality of pseudo-labels. Experimental results on three popular benchmarks showcase its superior performance compared to state-of-the-art methods.

cs.CV↗

arXiv2025

AlpaCare:Instruction-tuned Large Language Models for Medical Application

Instruction-finetuning (IFT) has become crucial in aligning Large Language Models (LLMs) with diverse human needs and has shown great potential in medical applications. However, previous studies mainly fine-tune LLMs on biomedical datasets with limited diversity, which often rely on benchmarks or narrow task scopes, and hence significantly limit the effectiveness on their medical instruction-following ability and generalizability. To bridge this gap, we propose creating a diverse, machine-generated medical IFT dataset, MedInstruct-52k, using GPT-4 and ChatGPT with a high-quality expert-curated seed set. We then fine-tune LLaMA-series models on the dataset to develop AlpaCare. Despite using a smaller domain-specific dataset than previous medical LLMs, AlpaCare not only demonstrates superior performance on medical applications, with up to 38.1% absolute gain over best baselines in medical free-form instruction evaluations, but also achieves 6.7% absolute gains averaged over multiple general domain benchmarks. Human evaluation further shows that AlpaCare consistently outperforms best baselines in terms of both correctness and helpfulness. We offer public access to our data, model, and codebase in https://github.com/XZhang97666/AlpaCare.

cs.CL↗

arXiv2025

Taste More, Taste Better: Diverse Data and Strong Model Boost Semi-Supervised Crowd Counting

Semi-supervised crowd counting is crucial for addressing the high annotation costs of densely populated scenes. Although several methods based on pseudo-labeling have been proposed, it remains challenging to effectively and accurately utilize unlabeled data. In this paper, we propose a novel framework called Taste More Taste Better (TMTB), which emphasizes both data and model aspects. Firstly, we explore a data augmentation technique well-suited for the crowd counting task. By inpainting the background regions, this technique can effectively enhance data diversity while preserving the fidelity of the entire scenes. Secondly, we introduce the Visual State Space Model as backbone to capture the global context information from crowd scenes, which is crucial for extremely crowded, low-light, and adverse weather scenarios. In addition to the traditional regression head for exact prediction, we employ an Anti-Noise classification head to provide less exact but more accurate supervision, since the regression head is sensitive to noise in manual annotations. We conduct extensive experiments on four benchmark datasets and show that our method outperforms state-of-the-art methods by a large margin. Code is publicly available on https://github.com/syhien/taste_more_taste_better.

cs.CV↗

arXiv2025

Steady Progress Beats Stagnation: Mutual Aid of Foundation and Conventional Models in Mixed Domain Semi-Supervised Medical Image Segmentation

Large pretrained visual foundation models exhibit impressive general capabilities. However, the extensive prior knowledge inherent in these models can sometimes be a double-edged sword when adapting them to downstream tasks in specific domains. In the context of semi-supervised medical image segmentation with domain shift, foundation models like MedSAM tend to make overconfident predictions, some of which are incorrect. The error accumulation hinders the effective utilization of unlabeled data and limits further improvements. In this paper, we introduce a Synergistic training framework for Foundation and Conventional models (SynFoC) to address the issue. We observe that a conventional model trained from scratch has the ability to correct the high-confidence mispredictions of the foundation model, while the foundation model can supervise it with high-quality pseudo-labels in the early training stages. Furthermore, to enhance the collaborative training effectiveness of both models and promote reliable convergence towards optimization, the consensus-divergence consistency regularization is proposed. We demonstrate the superiority of our method across four public multi-domain datasets. In particular, our method improves the Dice score by 10.31\% on the Prostate dataset. Our code is available at https://github.com/MQinghe/SynFoC .

cs.CV↗