Arrow Research search

Author name cluster

Han Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

51 papers
2 author rows

Possible papers

51

EAAI Journal 2026 Journal Article

Disentangling response sequences with causal invariance for knowledge tracing

  • Shengze Hu
  • Junjie Hu
  • Huali Yang
  • Jing Geng
  • Xinjia Ou
  • Zhuoran Xu
  • Han Wang
  • Tao Huang

To support personalized learning applications, Knowledge Tracing (KT) predicts students’ future performance by analyzing their historical interactions with questions. Despite advancements in neural network designs and integrated educational principles, existing KT methods often overlook data selection biases, leading to spurious correlations between future performance and response sequences. To address this, we propose a novel KT method based on Causal Invariance (CIKT), which robustly identifies causal relationships in response sequences. From a causal perspective, we develop a structural causal model for KT, guiding the design of CIKT’s four modules: a response sequence encoder, an attention-based causal identifier, a sequence causal intervener, and a future response predictor. The encoder extracts representations of future questions and response units. The causal identifier uses question-concept graphs and combines knowledge and temporal associations to estimate causal and trivial scores for each future question, effectively disentangling sequences into causal and trivial subsequences. Grounded in causal invariance, the intervener applies intervention operations to remove, replace, and invert elements within trivial subsequences, generating diverse sequences. The predictor then assigns the same performance prediction task to these intervened sequences to discover invariant causal relationships. Additionally, a question difficulty prediction task for trivial subsequences is introduced to prevent prediction shortcuts, reflecting group mastery levels. Extensive experiments on real-world datasets demonstrate CIKT’s superiority and good interpretability.

AAAI Conference 2026 Conference Paper

FreeAskWorld: An Interactive and Closed-Loop Simulator for Human-Centric Embodied AI

  • Yuhang Peng
  • Yizhou Pan
  • Xinning He
  • Jihaoyu Yang
  • Xinyu Yin
  • Han Wang
  • Xiaoji Zheng
  • Chao Gao

As embodied intelligence emerges as a core frontier in artificial intelligence research, simulation platforms must evolve beyond low-level physical interactions to capture complex, human-centered social behaviors. We introduce FreeAskWorld, an interactive simulation framework that integrates large language models (LLMs) for high-level behavior planning and semantically grounded interaction, informed by theories of intention and social cognition. Our framework supports scalable, realistic human-agent simulations and includes a modular data generation pipeline tailored for diverse embodied tasks.To validate the framework, we extend the classic Vision-and-Language Navigation (VLN) task into a semantically enriched Direction Inquiry setting, wherein agents can actively seek and interpret navigational guidance. We present and publicly release FreeAskWorld, a large-scale benchmark dataset comprising reconstructed environments, six diverse task types, 16 core object categories, 63,429 annotated sample frames, and more than 17 hours of interaction data to support training and evaluation of embodied AI systems. We benchmark VLN models, and human participants under both open-loop and closed-loop settings. Experimental results demonstrate that models fine-tuned on FreeAskWorld outperform their original counterparts, achieving enhanced semantic understanding and interaction competency. These findings underscore the efficacy of socially grounded simulation frameworks in advancing embodied AI systems toward sophisticated high-level planning and more naturalistic human-agent interaction.

JBHI Journal 2026 Journal Article

LELN: A Large Language Model-Dynamically Enhanced Learning Network for Patient Similarity Calculation

  • Zhichao Zhu
  • Bo Bai
  • Jianqiang Li
  • Han Wang
  • Rui Li
  • Lan Lan

The rapid expansion of Electronic Medical Record (EMR) data has advanced AI-driven patient similarity computation, a key technology for intelligent healthcare. However, the handling of heterogeneous EMR formats and the integration of domain knowledge constrain existing methods. While graph-based approaches show promise, they still struggle with these issues. To address this, we propose a Large Language Model-Dynamically Enhanced Learning Network (LELN), leveraging LLMs' commonsense knowledge and reasoning to dynamically structure EMR data and enhance medical knowledge integration. LELN in tegrates two LLM-basedmodules: DS-EE(DeepSeek-Event Extraction) extracts medical events to construct structured EMR event graphs, and DS-KB (DeepSeek-Knowledge Base) infers disease-relevant knowledge to augment feature representations. The model employs a dual-stage spatial-temporal feature aggregation strategy: a Graph Attention Network captures intra- and inter-event dependencies, followed by a Bidirectional Long-Short Term Memory (BiLSTM) with attention to model temporal disease progression. Additionally, a clinical prior-guided attention mechanism emphasizes discriminative diagnostic features, improving clinical relevance. Extensive experiments on heterogeneous datasets—a real-world Chinese dataset and public MIMIC-III—show LELN outperforms baselines, achieving F1 scores of 87. 66% and 85. 95%, demonstrating robustness and accuracy.

AAAI Conference 2026 Conference Paper

MEML-GRPO: Heterogeneous Multi-Expert Mutual Learning for RLVR Advancement

  • Weitao Jia
  • Jinghui Lu
  • Haiyang Yu
  • Siqi Wang
  • Guozhi Tang
  • An-Lan Wang
  • Weijie Yin
  • Dingkang Yang

Recent advances demonstrate that reinforcement learning with verifiable rewards (RLVR) significantly enhances the reasoning capabilities of large language models (LLMs). However, standard RLVR faces challenges with reward sparsity, where zero rewards from consistently incorrect candidate answers provide no learning signal, particularly in challenging tasks. To address this,we propose Multi-Expert Mutual Learning GRPO (MEML-GRPO), an innovative framework that utilizes diverse expert prompts as system prompts to generate a broader range of responses, substantially increasing the likelihood of identifying correct solutions. Additionally, we introduce an inter-expert mutual learning mechanism that facilitates knowledge sharing and transfer among experts, further boosting the model’s performance through RLVR. Extensive experiments across multiple reasoning benchmarks show that MEML-GRPO delivers significant improvements, achieving an average performance gain of 4.89% with Qwen and 11.33% with Llama, effectively overcoming the core limitations of traditional RLVR methods.

AAMAS Conference 2026 Conference Paper

Mind-Inference for Multi-Turn Emotional Support: Distinguishing Personal and Factual with Adaptive Memory

  • Peilin Li
  • Qiuyun Zhang
  • Yukun Lian
  • Chengxuan Li
  • Bin Guo
  • Mengqi Chen
  • Han Wang
  • Yuyang Fang

Emotional Support Conversation (ESC) requires models to track a user’s evolving psychological state and produce supportive responsesacrossmultipleturns. ExistingESCmethodsimproveempathy with emotion clues, strategy planning, or commonsense reasoning, but they often treat personal and factual information uniformly and therefore struggle with stale memories in long interactions. We present MIA, a compact framework for multi-turn ESC with three coordinated modules: EToM for hierarchical mind inference, PFD for adaptive Personal–Factual weighting, and ORM for removing obsolete psychological inferences during memory updates. Evaluations on ESConv and CPsyCounD show consistent gains over strong baselines, while ablations confirm that both adaptive factor weightingandobsolescence-awarememoryarenecessary. Theseresults indicate that explicit Personal/Factual separation is a practical design principle for robust multi-turn emotional support. ∗These authors contributed equally to this work. †Corresponding author. This work is licensed under a Creative Commons Attribution International 4. 0 License. Proc. of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026), C. Amato, L. Dennis, V. Mascardi, J. Thangarajah (eds.), May 25 – 29, 2026, Paphos, Cyprus. © 2026 International Foundation for Autonomous Agents and Multiagent Systems (www. ifaamas. org). https: //doi. org/10. 65109/CIMN4023

TMLR Journal 2026 Journal Article

Model-Free Learning with Heterogeneous Dynamical Systems: A Federated LQR Approach

  • Han Wang
  • Leonardo Felipe Toso
  • Aritra Mitra
  • James Anderson

We study a model-free federated linear quadratic regulator (LQR) problem where M agents with unknown, distinct yet similar dynamics collaboratively learn an optimal policy to minimize an average quadratic cost while keeping their data private. To exploit the similarity of the agents' dynamics, we propose to use federated learning (FL) to allow the agents to periodically communicate with a central server to train policies by leveraging a larger dataset from all the agents. With this setup, we seek to understand the following questions: (i) Is the learned common policy stabilizing for all agents? (ii) How close is the learned common policy to each agent's own optimal policy? (iii) Can each agent learn its own optimal policy faster by leveraging data from all agents? To answer these questions, we propose the federated and model-free algorithm FedLQR. Our analysis overcomes numerous technical challenges, such as heterogeneity in the agents’ dynamics, multiple local updates, and stability concerns. We show that FedLQR produces a common policy that, at each iteration, is stabilizing for all agents. Moreover, we prove that when learning each agent's optimal policy, FedLQR achieves a sample complexity reduction proportional to the number of agents M in a low-heterogeneity regime, compared to the single-agent setting.

AAAI Conference 2026 Conference Paper

Multi-Agent VLMs Guided Self-Training with PNU Loss for Low-Resource Offensive Content Detection

  • Han Wang
  • Deyi Ji
  • Junyu Lu
  • Lanyun Zhu
  • Hailong Zhang
  • Haiyang Wu
  • Liqun Liu
  • Peng Shu

Accurate detection of offensive content on social media demands high-quality labeled data; however, such data is often scarce due to the low prevalence of offensive instances and the high cost of manual annotation. To address this low-resource challenge, we propose a self-training framework that leverages abundant unlabeled data through collaborative pseudo-labeling. Starting with a lightweight classifier trained on limited labeled data, our method iteratively assigns pseudo-labels to unlabeled instances with the support of Multi-Agent Vision-Language Models (MA-VLMs). Unlabeled data on which the classifier and MA-VLMs agree are designated as the Agreed-Unknown set, while conflicting samples form the Disagreed-Unknown set. To enhance label reliability, MA-VLMs simulate dual perspectives, moderator and user, capturing both regulatory and subjective viewpoints. The classifier is optimized using a novel Positive-Negative-Unlabeled (PNU) loss, which jointly exploits labeled, Agreed-Unknown, and Disagreed-Unknown data while mitigating pseudo-label noise. Experiments on benchmark datasets demonstrate that our framework substantially outperforms baselines under limited supervision and approaches the performance of large-scale models.

AAAI Conference 2026 Conference Paper

Multimodal DeepResearcher: Generating Text-Chart Interleaved Reports from Scratch with Agentic Framework

  • Zhaorui Yang
  • Bo Pan
  • Han Wang
  • Yiyao Wang
  • Xingyu Liu
  • Luoxuan Weng
  • Yingchaojie Feng
  • Haozhe Feng

Visualizations play a crucial part in effective communication of concepts and information. Recent advances in reasoning and retrieval augmented generation have enabled Large Language Models (LLMs) to perform deep research and generate comprehensive reports. Despite its progress, existing deep research frameworks primarily focus on generating text-only content, leaving the automated generation of interleaved texts and visualizations underexplored. This novel task poses key challenges in designing informative visualizations and effectively integrating them with text reports. To address these challenges, we propose Formal Description of Visualization (FDV), a structured textual representation of charts that enables LLMs to learn from and generate diverse, high-quality visualizations. Building on this representation, we introduce Multimodal DeepResearcher, an agentic framework that decomposes the task into four stages: (1) researching, (2) exemplar report textualization, (3) planning and (4) multimodal report generation. For the evaluation of the generated reports, we develop MultimodalReportBench which contains 100 diverse topics as inputs, and a set of dedicated metrics for report and chart evaluation. Extensive experiments across models and evaluation methods demonstrate the effectiveness of Multimodal DeepResearcher. Notably, utilizing the same Claude 3.7 Sonnet model, Multimodal DeepResearcher achieves an 82% overall win rate over the baseline method.

AAAI Conference 2026 Conference Paper

R-Tuning: Wavelet-Decomposed Replay and Semantic Alignment for Continual Adaptation of Pretrained Time-Series Models

  • Tianyi Yin
  • Jingwei Wang
  • Chenze Wang
  • Han Wang
  • Jiexuan Cai
  • Min Liu
  • Yunlong Ma
  • Kun Gao

Pre-trained models have demonstrated exceptional generalization capabilities in time-series forecasting; however, adapting them to evolving data distributions remains a significant challenge. A key hurdle lies in accessing the original training data, as fine-tuning solely on new data often leads to catastrophic forgetting. To address this issue, we propose Replay Tuning (R-Tuning), a novel framework designed for the continual adaptation of pre-trained time-series models. R-Tuning constructs a unified latent space that captures both prior and current task knowledge through a frequency-aware replay strategy. Specifically, it augments model-generated samples via wavelet-based decomposition across multiple frequency bands, generating trend-preserving and fusion-enhanced variants to improve representation diversity and replay efficiency. To further reduce reliance on synthetic samples, R-Tuning introduces a latent consistency constraint that aligns new representations with the prior task space. This constraint guides joint optimization within a compact and semantically coherent latent space, ensuring robust knowledge retention and adaptation. Extensive experimental results demonstrate the superiority of R-Tuning, which reduces MAE and MSE by up to 46.9% and 46.8%, respectively, on new tasks, while preserving prior knowledge with gains of up to 5.7% and 6.0% on old tasks. Notably, under few-shot settings, R-Tuning outperforms all state-of-the-art baselines even when synthetic proxy samples account for only 5% of the new task dataset.

EAAI Journal 2025 Journal Article

A dynamic consensus-based decentralized method for multiple unmanned aerial vehicles cooperative target allocation

  • Han Wang
  • Xiaolong Liang
  • Jiaqiang Zhang
  • Aiwu Yang

This paper investigates multiple unmanned aerial vehicles dynamic target allocation problem with capability-requirement matching constraints. The objective is to achieve conflict-free allocation through distributed collaboration under a dynamic environment. To address this problem, we propose a decentralized method named Dynamic Consensus-Based Group Algorithm (DCBGA). This algorithm utilizes a rule-based method known as State-Event-Condition-Action (SECA) as the decision-making framework. In addition, this algorithm consists of iterations between two phases: target selection and conflict resolution. In the target selection phase, each unmanned aerial vehicle (UAV) selects targets in a market-based method and follows the principle of “capability-requirement matching”. In the conflict resolution phase, a consensus-based mechanism is designed so that each UAV communicates with its neighbors to generate a conflict-free allocation result. In the simulation section, this paper performs convergence analysis, sensitivity analysis, and optimization comparisons. The simulation results confirm both the convergence and robustness of the DCBGA. Furthermore, the comparison results demonstrate that the proposed algorithm exhibits better optimization performance than other advanced algorithms.

AAAI Conference 2025 Conference Paper

Apollo-Forecast: Overcoming Aliasing and Inference Speed Challenges in Language Models for Time Series Forecasting

  • Tianyi Yin
  • Jingwei Wang
  • Yunlong Ma
  • Han Wang
  • Chenze Wang
  • Yukai Zhao
  • Min Liu
  • Weiming Shen

Encoding time series into tokens and using language models for processing has been shown to substantially augment the models' ability to generalize to unseen tasks. However, existing language models for time series forecasting encounter several obstacles, including aliasing distortion and prolonged inference times, primarily due to the limitations of quantization processes and the computational demands of large models. This paper introduces Apollo-Forecast, a novel framework that tackles these challenges with two key innovations: the Anti-Aliasing Quantization Module (AAQM) and the Race Decoding (RD) technique. AAQM adeptly encodes sequences into tokens while mitigating high-frequency noise in the original signals, thus enhancing both signal fidelity and overall quantization efficiency. RD employs a draft model to enable parallel processing and results integration, which markedly accelerates the inference speed for long-term predictions, particularly in large-scale models. Extensive experiments on various real-world datasets show that Apollo-Forecast outperforms state-of-the-art methods by 35.41% and 18.99% in WQL and MASE metrics, respectively, in zero-shot scenarios. Furthermore, our method achieves an acceleration of 1.9X-2.7X in inference speed over the baseline methods.

NeurIPS Conference 2025 Conference Paper

Conformal Prediction in The Loop: A Feedback-Based Uncertainty Model for Trajectory Optimization

  • Han Wang
  • Chao Ning

Conformal Prediction (CP) is a powerful statistical machine learning tool to construct uncertainty sets with coverage guarantees, which has fueled its extensive adoption in generating prediction regions for decision-making tasks, e. g. , Trajectory Optimization (TO) in uncertain environments. However, existing methods predominantly employ a sequential scheme, where decisions rely unidirectionally on the prediction regions, and consequently the information from decision-making fails to be fed back to instruct CP. In this paper, we propose a novel Feedback-Based CP (Fb-CP) framework for shrinking-horizon TO with a joint risk constraint over the entire mission time. Specifically, a CP-based posterior risk calculation method is developed by fully leveraging the realized trajectories to adjust the posterior allowable risk, which is then allocated to future times to update prediction regions. In this way, the information in the realized trajectories is continuously fed back to the CP, enabling attractive feedback-based adjustments of the prediction regions and a provable online improvement in trajectory performance. Furthermore, we theoretically prove that such adjustments consistently maintain the coverage guarantees of the prediction regions, thereby ensuring provable safety. Additionally, we develop a decision-focused iterative risk allocation algorithm with theoretical convergence analysis for allocating the posterior allowable risk which closely aligns with Fb-CP. Furthermore, we extend the proposed method to handle distribution shift. The effectiveness and superiority of the proposed method are demonstrated through benchmark experiments.

JBHI Journal 2025 Journal Article

Edge-Guided Multi-Scale Frequency Attention Network for Gastrointestinal Cancer Image Segmentation

  • Zhiwen Liao
  • Qi Wang
  • Xinyi Tang
  • Han Wang
  • Jun Hu
  • Pengxiang Su
  • Evangelos K. Markakis
  • Peng Luo

Image segmentation is a critical technology to improve the accuracy of clinical decisions and treatments in computer-aided diagnostic systems. However, the diverse morphology and fuzzy boundaries of gastrointestinal tumors incur substantial challenges for existing segmentation models, leading to inaccurate feature capture and generating suboptimal results. For solving these problems, we design an edge-guided multi-scale frequency attention network for the gastrointestinal tumor segmentation task, termed EGMFA-Net, which consists of a Kernel Adaptive Enhancement Module (KAEM) and a Frequency-domain Self-attention Module (FDSA). Specifically, KAEM adaptively adjusts the feature extraction kernel based on the morphology of different lesion regions, which enhances the recognition of different morphology regions via a progressive optimization strategy of feature expression. Furthermore, FDSA effectively aggregates multi-scale features in the frequency domain to achieve global receptive fields while preserving more high-frequency details, thereby enhancing adaptability to complex pathological contexts. Extensive experiments on eight medical image benchmark datasets, including SEED, Kvasir, ClinicDB, ColonDB, ETIS, BKAI, CVC-300, and Synapse, show that EGMFA-Net attains state-of-the-art performance over existing methods. Our implementation is available at https://github.com/med-segment/egmfa-net.

NeurIPS Conference 2025 Conference Paper

Embodied Cognition Augmented End2End Autonomous Driving

  • Ling Niu
  • Xiaoji Zheng
  • Han Wang
  • Ziyuan Yang
  • Chen Zheng
  • Bokui Chen
  • Jiangtao Gong

In recent years, vision-based end-to-end autonomous driving has emerged as a new paradigm. However, popular end-to-end approaches typically rely on visual feature extraction networks trained under label supervision. This limited supervision framework restricts the generality and applicability of driving models. In this paper, we propose a novel paradigm termed $E^{3}AD$, which advocates for comparative learning between visual feature extraction networks and the general EEG large model, in order to learn latent human driving cognition for enhancing end-to-end planning. In this work, we collected a cognitive dataset for the mentioned contrastive learning process. Subsequently, we investigated the methods and potential mechanisms for enhancing end-to-end planning with human driving cognition, using popular driving models as baselines on publicly available autonomous driving datasets. Both open-loop and closed-loop tests are conducted for a comprehensive evaluation of planning performance. Experimental results demonstrate that the $E^{3}AD$ paradigm significantly enhances the end-to-end planning performance of baseline models. Ablation studies further validate the contribution of driving cognition and the effectiveness of comparative learning process. To the best of our knowledge, this is the first work to integrate human driving cognition for improving end-to-end autonomous driving planning. It represents an initial attempt to incorporate embodied cognitive data into end-to-end autonomous driving, providing valuable insights for future brain-inspired autonomous driving systems. Our code will be made available at https: //github. com/AIR-DISCOVER/E-cubed-AD.

ICLR Conference 2025 Conference Paper

GLOMA: Global Video Text Spotting with Morphological Association

  • Han Wang
  • Yanjie Wang
  • Yang Li
  • Can Huang

Video Text Spotting (VTS) is a fundamental visual task that aims to predict the trajectories and content of texts in a video. Previous works usually conduct local associations and apply IoU-based distance and complex post-processing procedures to boost performance, ignoring the abundant temporal information and the morphological characteristics in VTS. In this paper, we propose \model{} to model the tracking problem as global associations and utilize the Gaussian Wasserstein distance to guide the morphological correlation between frames. Our main contributions can be summarized as three folds. 1). We propose a Transformer-based global tracking method \model{} for VTS and associate multiple frames simultaneously. 2). We introduce a Wasserstein distance-based method to conduct positional associations between frames. 3). We conduct extensive experiments on public datasets. On the ICDAR2015 video dataset, \model{} achieves \textbf{56.0} MOTA with \textbf{4.6} absolute improvement compared with the previous SOTA method and outperforms the previous Transformer-based method by a significant \textbf{8.3} MOTA.

ICLR Conference 2025 Conference Paper

Innovative Thinking, Infinite Humor: Humor Research of Large Language Models through Structured Thought Leaps

  • Han Wang
  • Yilin Zhao
  • Dian Li
  • Xiaohan Wang
  • Sinbadliu
  • Xuguang Lan
  • Hui Wang

Humor is previously regarded as a gift exclusive to humans for the following reasons. Humor is a culturally nuanced aspect of human language, presenting challenges for its understanding and generation. Humor generation necessitates a multi-hop reasoning process, with each hop founded on proper rationales. Although many studies, such as those related to GPT-o1, focus on logical reasoning with reflection and correction, they still fall short in humor generation. Due to the sparsity of the knowledge graph in creative thinking, it is arduous to achieve multi-hop reasoning. Consequently, in this paper, we propose a more robust framework for addressing the humor reasoning task, named LoL. LoL aims to inject external information to mitigate the sparsity of the knowledge graph, thereby enabling multi-hop reasoning. In the first stage of LoL, we put forward an automatic instruction-evolution method to incorporate the deeper and broader thinking processes underlying humor. Judgment-oriented instructions are devised to enhance the model's judgment capability, dynamically supplementing and updating the sparse knowledge graph. Subsequently, through reinforcement learning, the reasoning logic for each online-generated response is extracted using GPT-4o. In this process, external knowledge is re-introduced to aid the model in logical reasoning and the learning of human preferences. Finally, experimental results indicate that the combination of these two processes can enhance both the model's judgment ability and its generative capacity. These findings deepen our comprehension of the creative capabilities of large language models (LLMs) and offer approaches to boost LLMs' creative abilities for cross-domain innovative applications.

ICLR Conference 2025 Conference Paper

Learning to Communicate Through Implicit Communication Channels

  • Han Wang
  • Binbin Chen 0005
  • Tieying Zhang
  • Baoxiang Wang 0001

Effective communication is an essential component in collaborative multi-agent systems. Situations where explicit messaging is not feasible have been common in human society throughout history, which motivate the study of implicit communication. Previous works on learning implicit communication mostly rely on theory of mind (ToM), where agents infer the mental states and intentions of others by interpreting their actions. However, ToM-based methods become less effective in making accurate inferences in complex tasks. In this work, we propose the Implicit Channel Protocol (ICP) framework, which allows agents to communicate through implicit communication channels similar to the explicit ones. ICP leverages a subset of actions, denoted as the scouting actions, and a mapping between information and these scouting actions that encodes and decodes the messages. We propose training algorithms for agents to message and act, including learning with a randomly initialized information map and with a delayed information map. The efficacy of ICP has been tested on the tasks of Guessing Numbers, Revealing Goals, and Hanabi, where ICP significantly outperforms baseline methods through more efficient information transmission.

ICML Conference 2025 Conference Paper

MimicMotion: High-Quality Human Motion Video Generation with Confidence-aware Pose Guidance

  • Yuang Zhang
  • Jiaxi Gu
  • Li-Wen Wang
  • Han Wang
  • Junqi Cheng
  • Yuefeng Zhu
  • Fangyuan Zou

In recent years, while generative AI has advanced significantly in image generation, video generation continues to face challenges in controllability, length, and detail quality, which hinder its application. We present MimicMotion, a framework for generating high-quality human videos of arbitrary length using motion guidance. Our approach has several highlights. Firstly, we introduce confidence-aware pose guidance that ensures high frame quality and temporal smoothness. Secondly, we introduce regional loss amplification based on pose confidence, which reduces image distortion in key regions. Lastly, we propose a progressive latent fusion strategy to generate long and smooth videos. Experiments demonstrate the effectiveness of our approach in producing high-quality human motion videos. Videos and comparisons are available at https: //tencent. github. io/MimicMotion.

NeurIPS Conference 2025 Conference Paper

MRSAudio: A Large-Scale Multimodal Recorded Spatial Audio Dataset with Refined Annotations

  • Wenxiang Guo
  • Changhao Pan
  • Zhiyuan Zhu
  • Xintong Hu
  • Yu Zhang
  • Li Tang
  • Rui Yang
  • Han Wang

Humans rely on multisensory integration to perceive spatial environments, where auditory cues enable sound source localization in three-dimensional space. Despite the critical role of spatial audio in immersive technologies such as VR/AR, most existing multimodal datasets provide only monaural audio, which limits the development of spatial audio generation and understanding. To address these challenges, we introduce MRSAudio, a large-scale multimodal spatial audio dataset designed to advance research in spatial audio understanding and generation. MRSAudio spans four distinct components: MRSLife, MRSSpeech, MRSMusic, and MRSSing, covering diverse real-world scenarios. The dataset includes synchronized binaural and ambisonic audio, exocentric and egocentric video, motion trajectories, and fine-grained annotations such as transcripts, phoneme boundaries, lyrics, scores, and prompts. To demonstrate the utility and versatility of MRSAudio, we establish five foundational tasks: audio spatialization, and spatial text to speech, spatial singing voice synthesis, spatial music generation and sound event localization and detection. Results show that MRSAudio enables high-quality spatial modeling and supports a broad range of spatial audio research. Demos and dataset access are available at https: //mrsaudio. github. io.

ICML Conference 2025 Conference Paper

Off-Policy Evaluation under Nonignorable Missing Data

  • Han Wang
  • Yang Xu 0089
  • Wenbin Lu
  • Rui Song 0006

Off-Policy Evaluation (OPE) aims to estimate the value of a target policy using offline data collected from potentially different policies. In real-world applications, however, logged data often suffers from missingness. While OPE has been extensively studied in the literature, a theoretical understanding of how missing data affects OPE results remains unclear. In this paper, we investigate OPE in the presence of monotone missingness and theoretically demonstrate that the value estimates remain unbiased under ignorable missingness but can be biased under nonignorable (informative) missingness. To retain the consistency of value estimation, we propose an inverse probability weighting value estimator and conduct statistical inference to quantify the uncertainty of the estimates. Through a series of numerical experiments, we empirically demonstrate that our proposed estimator yields a more reliable value inference under missing data.

ECAI Conference 2025 Conference Paper

Overexposed Frame Reconstruction in Ultra-High-Speed Imaging via Event-Guided Diffusion Models

  • Han Wang
  • Sijia Liu
  • Juntao Wu
  • Xirui Zhang
  • Zhou Wang
  • Xiaofeng Yang
  • Yaoxiong Wang
  • Saiao Zhou

Ultra-high-speed cameras frequently suffer from severe overexposure in scenarios involving extreme brightness transitions, significantly degrading image quality and obscuring critical visual details. To address this issue, we propose a novel reconstruction method combining neuromorphic sensors with state-of-the-art diffusion models. Our approach leverages the asynchronous, high-temporal-resolution, and high-dynamic-range capabilities of neuro-morphic sensors to capture rapid brightness variations, subsequently utilizing conditional diffusion models to reconstruct high-quality frames from sparse event data. We validated the proposed method through experiments conducted under three challenging lighting conditions. The results demonstrate that our approach effectively recovers detailed visual content in severely overexposed frames, significantly outperforming traditional frame-based imaging techniques.

NeurIPS Conference 2025 Conference Paper

PointMapPolicy: Structured Point Cloud Processing for Multi-Modal Imitation Learning

  • Xiaogang Jia
  • Qian Wang
  • Anrui Wang
  • Han Wang
  • Balázs Gyenes
  • Emiliyan Gospodinov
  • Xinkai Jiang
  • Ge Li

Robotic manipulation systems benefit from complementary sensing modalities, where each provides unique environmental information. Point clouds capture detailed geometric structure, while RGB images provide rich semantic context. Current point cloud methods struggle to capture fine-grained detail, especially for complex tasks, which RGB methods lack geometric awareness, which hinders their precision and generalization. We introduce PointMapPolicy, a novel approach that conditions diffusion policies on structured grids of points without downsampling. The resulting data type makes it easier to extract shape and spatial relationships from observations, and can be transformed between reference frames. Yet due to their structure in a regular grid, we enable the use of established computer vision techniques directly to 3D data. Using xLSTM as a backbone, our model efficiently fuses the point maps with RGB data for enhanced multi-modal perception. Through extensive experiments on the RoboCasa and CALVIN benchmarks and real robot evaluations, we demonstrate that our method achieves state-of-the-art performance across diverse manipulation tasks. The overview and demos are available on our project page: https: //point-map. github. io/Point-Map/

JBHI Journal 2025 Journal Article

SSGraphDTI: A Drug-Target Interaction Prediction Method Integrated Structural and Dynamic Systemic Biology Attributes

  • Haotian Guan
  • Tian Bai
  • Jingtong Zhao
  • Wenhao Li
  • Han Wang

Drug-Target Interaction (DTI) is a crucial aspect of pharmaceutical development. However, biochemical experiments are prohibitively expensive to identify these interactions on a large scale, while the computational approach is still on the way to making a highly reliable prediction. For the purpose of promoting prediction accuracy, drug-related molecular networks are gradually introduced to this task to furnish valuable information. We hypothesized that integrating structural and systemic biological attributes could effectively enhance the performance of DTI prediction and proposed a novel DTI prediction model, SSGraphDTI, which integrated two aforementioned attributes. Specifically, the structural attributes of drugs and targets are extracted using independent convolutional neural network based models from the Simplified Molecular Input Line Entry System of drugs and the amino acid sequences of targets, respectively. Meanwhile, the systemic biological attributes of drug-target pairs are obtained through graph representation learning on the dynamically constructed heterogeneous drug-target interaction network. SSGraphDTI was meticulously trained and rigorously tested on the benchmark Dataset_DrugBank, achieving an improvement of approximately 1. 0% across five metrics compared to recent comparable methods. These results underscore the potential of combining both structural and systemic information for accurate DTI prediction. Benefiting from the fact that the input consists solely of structural data without requiring interaction information, the model effectively addresses the “cold-start problem” in drug discovery. Furthermore, by extracting systemic attributes directly from the dynamically constructed DTI networks, the model maintains strong predictive performance even when data is limited. The source code is available at https://github.com/NENUBioCompute/SSGraphDTI.

EAAI Journal 2024 Journal Article

A novel CT image segmentation model

  • Jingdong Yang
  • Han Wang
  • Wei Liu
  • Xianyou Zheng
  • Xiaolin Zhang
  • Shaoqing Yu

Convolutional neural networks (CNNs) can be used for clinical medical image segmentation to improve detection efficiency and accuracy. However, because of the fixed size of convolution kernel and receptive field for existing CNN model, some associated pixel features are ignored and segmentation performance is impaired. Therefore, we propose an effective image segmentation model, COPANet, which uses RepVGG as the backbone and replaces the standard convolution with dilated convolution of multiple kernels to increase the receptive field size, so that COPANet can make full use of pixel correlations and acquire contexture semantic information. We also build skip-connections between global and local features, where a parallel attention (PANet) is employed to extract important location information in the downsampling. PANet is a fused mechanism integrating channel attention with spatial attention in parallel, which can fully extract more global information. We also apply weighted combined loss to reduce the effect of class imbalance of foreground and background pixels on segmentation performance and speed up convergence. In addition, we conduct experiments on 480 cases of clinical CT sinus from Shanghai Tongji Hospital and 236 cases of CT patella fracture from Shanghai Sixth People's Hospital. The evaluation indexes after 5-Fold cross-validation are as follows: Precision is 92. 76% and 96. 55%, Recall is 88. 25% and 57. 58%, Specificity is 99. 85% and 96. 55%, IOU is 91. 08% and 77. 38%, Hausdorff Distance is 2. 8135 and 11. 1879, respectively. Compared with the state-of-the-art models, COPANet has higher segmentation accuracy and better generalization performance, which can assist the clinical diagnosis.

RLJ Journal 2024 Journal Article

A Simple Mixture Policy Parameterization for Improving Sample Efficiency of CVaR Optimization

  • Yudong Luo
  • Yangchen Pan
  • Han Wang
  • Philip Torr
  • Pascal Poupart

Reinforcement learning algorithms utilizing policy gradients (PG) to optimize Conditional Value at Risk (CVaR) face significant challenges with sample inefficiency, hindering their practical applications. This inefficiency stems from two main facts: a focus on tail-end performance that overlooks many sampled trajectories, and the potential of gradient vanishing when the lower tail of the return distribution is overly flat. To address these challenges, we propose a simple mixture policy parameterization. This method integrates a risk-neutral policy with an adjustable policy to form a risk-averse policy. By employing this strategy, all collected trajectories can be utilized for policy updating, and the issue of vanishing gradients is counteracted by stimulating higher returns through the risk-neutral component, thus lifting the tail and preventing flatness. Our empirical study reveals that this mixture parameterization is uniquely effective across a variety of benchmark domains. Specifically, it excels in identifying risk-averse CVaR policies in some Mujoco environments where the traditional CVaR-PG fails to learn a reasonable policy.

RLC Conference 2024 Conference Paper

A Simple Mixture Policy Parameterization for Improving Sample Efficiency of CVaR Optimization

  • Yudong Luo
  • Yangchen Pan
  • Han Wang
  • Philip Torr
  • Pascal Poupart

Reinforcement learning algorithms utilizing policy gradients (PG) to optimize Conditional Value at Risk (CVaR) face significant challenges with sample inefficiency, hindering their practical applications. This inefficiency stems from two main facts: a focus on tail-end performance that overlooks many sampled trajectories, and the potential of gradient vanishing when the lower tail of the return distribution is overly flat. To address these challenges, we propose a simple mixture policy parameterization. This method integrates a risk-neutral policy with an adjustable policy to form a risk-averse policy. By employing this strategy, all collected trajectories can be utilized for policy updating, and the issue of vanishing gradients is counteracted by stimulating higher returns through the risk-neutral component, thus lifting the tail and preventing flatness. Our empirical study reveals that this mixture parameterization is uniquely effective across a variety of benchmark domains. Specifically, it excels in identifying risk-averse CVaR policies in some Mujoco environments where the traditional CVaR-PG fails to learn a reasonable policy.

NeurIPS Conference 2024 Conference Paper

ALI-Agent: Assessing LLMs' Alignment with Human Values via Agent-based Evaluation

  • jingnan zheng
  • Han Wang
  • An Zhang
  • Tai D. Nguyen
  • Jun Sun
  • Tat-Seng Chua

Large Language Models (LLMs) can elicit unintended and even harmful content when misaligned with human values, posing severe risks to users and society. To mitigate these risks, current evaluation benchmarks predominantly employ expert-designed contextual scenarios to assess how well LLMs align with human values. However, the labor-intensive nature of these benchmarks limits their test scope, hindering their ability to generalize to the extensive variety of open-world use cases and identify rare but crucial long-tail risks. Additionally, these static tests fail to adapt to the rapid evolution of LLMs, making it hard to evaluate timely alignment issues. To address these challenges, we propose ALI-Agent, an evaluation framework that leverages the autonomous abilities of LLM-powered agents to conduct in-depth and adaptive alignment assessments. ALI-Agent operates through two principal stages: Emulation and Refinement. During the Emulation stage, ALI-Agent automates the generation of realistic test scenarios. In the Refinement stage, it iteratively refines the scenarios to probe long-tail risks. Specifically, ALI-Agent incorporates a memory module to guide test scenario generation, a tool-using module to reduce human labor in tasks such as evaluating feedback from target LLMs, and an action module to refine tests. Extensive experiments across three aspects of human values--stereotypes, morality, and legality--demonstrate that ALI-Agent, as a general evaluation framework, effectively identifies model misalignment. Systematic analysis also validates that the generated test scenarios represent meaningful use cases, as well as integrate enhanced measures to probe long-tail risks.

JBHI Journal 2024 Journal Article

Application of Zero-Watermarking Scheme Based on Swin Transformer for Securing the Metaverse Healthcare Data

  • Baoru Han
  • Han Wang
  • Dawei Qiao
  • Jia Xu
  • Tianyu Yan

The existing medical image privacy solutions cannot completely solve the security problems created by applying the metaverse healthcare system. A robust zero-watermarking scheme based on the Swin Transformer is proposed in this article to improve the security of medical images in the metaverse healthcare system. This scheme uses a pretrained Swin Transformer to extract deep features from the original medical images with a good generalization performance and multiscale, and binary feature vectors are generated by using the mean hashing algorithm. Then, the logistic chaotic encryption algorithm boosts the security of the watermarking image by encrypting it. Finally, an encrypted watermarking image is XORed with the binary feature vector to create a zero-watermarking, and the validity of the proposed scheme is verified through experimentation. According to the results of the experiments, the proposed scheme has excellent robustness to common attacks and geometric attacks, and implements privacy protections for medical image security transmissions in the metaverse. The research results provide a reference for the data security and privacy protection of the metaverse healthcare system.

NeurIPS Conference 2024 Conference Paper

Bridging OOD Detection and Generalization: A Graph-Theoretic View

  • Han Wang
  • Yixuan Li

In the context of modern machine learning, models deployed in real-world scenarios often encounter diverse data shifts like covariate and semantic shifts, leading to challenges in both out-of-distribution (OOD) generalization and detection. Despite considerable attention to these issues separately, a unified framework for theoretical understanding and practical usage is lacking. To bridge the gap, we introduce a graph-theoretic framework to jointly tackle both OOD generalization and detection problems. By leveraging the graph formulation, data representations are obtained through the factorization of the graph's adjacency matrix, enabling us to derive provable error quantifying OOD generalization and detection performance. Empirical results showcase competitive performance in comparison to existing methods, thereby validating our theoretical underpinnings.

IJCAI Conference 2024 Conference Paper

Carbon Market Simulation with Adaptive Mechanism Design

  • Han Wang
  • Wenhao Li
  • Hongyuan Zha
  • Baoxiang Wang

A carbon market is a market-based tool that incentivizes economic agents to align individual profits with the global utility, i. e. , reducing carbon emissions to tackle climate change. Cap and trade stands as a critical principle based on allocating and trading carbon allowances (carbon emission credit), enabling economic agents to follow planned emissions and penalizing excess emissions. A central authority is responsible for introducing and allocating those allowances in cap and trade. However, the complexity of carbon market dynamics makes accurate simulation intractable, which in turn hinders the design of effective allocation strategies. To address this, we propose an adaptive mechanism design framework, simulating the market using hierarchical, model-free multi-agent reinforcement learning (MARL). Government agents allocate carbon credits, while enterprises engage in economic activities and carbon trading. This framework illustrates agents’ behavior comprehensively. Numerical results show MARL enables government agents to balance productivity, equality, and carbon emissions. Our project is available at https: //anonymous. 4open. science/r/Carbon-Simulator.

NeurIPS Conference 2024 Conference Paper

Exploiting the Replay Memory Before Exploring the Environment: Enhancing Reinforcement Learning Through Empirical MDP Iteration

  • Hongming Zhang
  • Chenjun Xiao
  • Chao Gao
  • Han Wang
  • Bo Xu
  • Martin Müller

Reinforcement learning (RL) algorithms are typically based on optimizing a Markov Decision Process (MDP) using the optimal Bellman equation. Recent studies have revealed that focusing the optimization of Bellman equations solely on in-sample actions tends to result in more stable optimization, especially in the presence of function approximation. Upon on these findings, in this paper, we propose an Empirical MDP Iteration (EMIT) framework. EMIT constructs a sequence of empirical MDPs using data from the growing replay memory. For each of these empirical MDPs, it learns an estimated Q-function denoted as $\widehat{Q}$. The key strength is that by restricting the Bellman update to in-sample bootstrapping, each empirical MDP converges to a unique optimal $\widehat{Q}$ function. Furthermore, gradually expanding from the empirical MDPs to the original MDP induces a monotonic policy improvement. Instead of creating entirely new algorithms, we demonstrate that EMIT can be seamlessly integrated with existing online RL algorithms, effectively acting as a regularizer for contemporary Q-learning methods. We show this by implementing EMIT for two representative RL algorithms, DQN and TD3. Experimental results on Atari and MuJoCo benchmarks show that EMIT significantly reduces estimation errors and substantially improves the performance of both algorithms.

TMLR Journal 2024 Journal Article

Federated TD Learning with Linear Function Approximation under Environmental Heterogeneity

  • Han Wang
  • Aritra Mitra
  • Hamed Hassani
  • George J. Pappas
  • James Anderson

We initiate the study of federated reinforcement learning under environmental heterogeneity by considering a policy evaluation problem. Our setup involves $N$ agents interacting with environments that share the same state and action space but differ in their reward functions and state transition kernels. Assuming agents can communicate via a central server, we ask: \textit{Does exchanging information expedite the process of evaluating a common policy?} To answer this question, we provide the first comprehensive finite-time analysis of a federated temporal difference (TD) learning algorithm with linear function approximation, while accounting for Markovian sampling, heterogeneity in the agents' environments, and multiple local updates to save communication. Our analysis crucially relies on several novel ingredients: (i) deriving perturbation bounds on TD fixed points as a function of the heterogeneity in the agents' underlying Markov decision processes (MDPs); (ii) introducing a virtual MDP to closely approximate the dynamics of the federated TD algorithm; and (iii) using the virtual MDP to make explicit connections to federated optimization. Putting these pieces together, we prove that in a low-heterogeneity regime, exchanging model estimates leads to linear convergence speedups in the number of agents. Our theoretical contribution is significant in that it is the first result of its kind in multi-agent/federated reinforcement learning that complements the numerous analogous results in heterogeneous federated optimization.

ICML Conference 2024 Conference Paper

Few-Shot Character Understanding in Movies as an Assessment to Meta-Learning of Theory-of-Mind

  • Mo Yu
  • Qiujing Wang
  • Shunchi Zhang
  • Yisi Sang
  • Kangsheng Pu
  • Zekai Wei
  • Han Wang
  • Liyan Xu

When reading a story, humans can quickly understand new fictional characters with a few observations, mainly by drawing analogies to fictional and real people they already know. This reflects the few-shot and meta-learning essence of humans’ inference of characters’ mental states, i. e. , theory-of-mind (ToM), which is largely ignored in existing research. We fill this gap with a novel NLP dataset in a realistic narrative understanding scenario, ToM-in-AMC. Our dataset consists of $\sim$1, 000 parsed movie scripts, each corresponding to a few-shot character understanding task that requires models to mimic humans’ ability of fast digesting characters with a few starting scenes in a new movie. We further propose a novel ToM prompting approach designed to explicitly assess the influence of multiple ToM dimensions. It surpasses existing baseline models, underscoring the significance of modeling multiple ToM dimensions for our task. Our extensive human study verifies that humans are capable of solving our problem by inferring characters’ mental states based on their previously seen movies. In comparison, all the AI systems lag $>20%$ behind humans, highlighting a notable limitation in existing approaches’ ToM capabilities. Code and data are available at https: //github. com/ShunchiZhang/ToM-in-AMC

ICRA Conference 2024 Conference Paper

Grasp Manipulation Relationship Detection based on Graph Sample and Aggregation

  • Jiayuan Luo
  • Yaxin Liu
  • Han Wang
  • Mengyuan Ding
  • Xuguang Lan

In multi-object stacking scenarios, exploring the relationships among objects and determining the correct sequence of operations are crucial for robotic manipulation. However, previous algorithms inefficiently combine global and local information, often focusing solely on the local features of objects or the interactions of object features at a global level. This approach leads to imbalanced distribution of features and the generation of redundant or missing relationships in complex scenes, such as multi-object stacking and partial occlusion. To address this issue, we have developed a grasp manipulation relationship detection algorithm called Graph Sampling Aggregation Network for Visual Manipulation Relationship Detection (GSAGED). This algorithm assists robots in detecting targets in complex scenes and determining the appropriate grasping order. Firstly, the Positional Encoding Module in GSAGED enhances object feature information by considering global contexts. Secondly, the Graph Sampling Aggregation method effectively integrates global and local information, relieving imbalanced distribution of features. Finally, we applied the developed algorithm to a physical robot for grasping. Experimental results on the Visual Manipulation Relationship Dataset (VMRD) and the large-scale relational grasp dataset named REGRAD demonstrate that our method significantly improves the accuracy of relationship detection in complex scenes and exhibits robust generalization capabilities in real-world applications.

AIJ Journal 2024 Journal Article

Investigating the properties of neural network representations in reinforcement learning

  • Han Wang
  • Erfan Miahi
  • Martha White
  • Marlos C. Machado
  • Zaheer Abbas
  • Raksha Kumaraswamy
  • Vincent Liu
  • Adam White

In this paper we investigate the properties of representations learned by deep reinforcement learning systems. Much of the early work on representations for reinforcement learning focused on designing fixed-basis architectures to achieve properties thought to be desirable, such as orthogonality and sparsity. In contrast, the idea behind deep reinforcement learning methods is that the agent designer should not encode representational properties, but rather that the data stream should determine the properties of the representation—good representations emerge under appropriate training schemes. In this paper we bring these two perspectives together, empirically investigating the properties of representations that support transfer in reinforcement learning. We introduce and measure six representational properties over more than 25, 000 agent-task settings. We consider Deep Q-learning agents with different auxiliary losses in a pixel-based navigation environment, with source and transfer tasks corresponding to different goal locations. We develop a method to better understand why some representations work better for transfer, through a systematic approach varying task similarity and measuring and correlating representation properties with transfer performance. We demonstrate the generality of the methodology by investigating representations learned by a Rainbow agent that successfully transfers across Atari 2600 game modes.

TMLR Journal 2024 Journal Article

Offline Reinforcement Learning via Tsallis Regularization

  • Lingwei Zhu
  • Matthew Kyle Schlegel
  • Han Wang
  • Martha White

Offline reinforcement learning (RL) focuses on learning a good policy from a fixed dataset. The dataset is generated by an unknown behavior policy through interactions with the environment and contains only a subset of the state-action spaces. Standard off-policy algorithms often perform poorly in this setting, suffering from errorneously optimistic values incurred by the out-of-distribution (OOD) actions not present in the dataset. The optimisim cannot be corrected as no further interaction with the environment is possible. Imposing divergence regularization and in-sample constraints are among the most popular methods to overcoming the issue by ensuring that the learned policy stays close to the behavior policy to minimize the occurrence of OOD actions. This paper proposes Tsallis regularization for offline RL, which aligns the induced sparsemax policies to the in-sample constraint. Sparsemax interpolates existing methods utilizing hard-max and softmax policies, in that only a subset of actions contributes non-zero action probability as compared to softmax (all actions) and hard-max (single action). We leverage this property to model the behavior policy and show that under several assumptions the learned sparsemax policies may have sparsity-conditional KL divergence to the behavior policy, making Tsallis regularization especially suitable for the Behavior Cloning methods. We propose a novel actor-critic algorithm: Tsallis Advantage Weighted Actor-Critic (Tsallis AWAC) generalizing AWAC and analyze its performance in standard Mujoco environments. Our code is available at \url{https://github.com/lingweizhu/tsallis_regularization}.

IROS Conference 2024 Conference Paper

Real-Time Semantic Segmentation in Natural Environments with SAM-assisted Sim-to-Real Domain Transfer

  • Han Wang
  • Ruben Mascaro
  • Margarita Chli
  • Lucas Teixeira

Semantic segmentation plays a pivotal role in many robotic applications requiring high-level scene understanding, such as smart farming, where the precise identification of trees or plants can aid navigation and crop monitoring tasks. While deep-learning-based semantic segmentation approaches have reached outstanding performance in recent years, they demand large amounts of labeled data for training. Inspired by modern Unsupervised Domain Adaptation (UDA) techniques, in this paper, we introduce a two-step training pipeline specifically tailored to challenging natural scenes, where the availability of annotated data is often quite limited. Our strategy involves the initial training of a powerful domain adaptive architecture, followed by a refinement stage, where segmentation masks predicted by the Segment Anything Model (SAM) are used to improve the accuracy of the predictions on the target dataset. These refined predictions serve as pseudo-labels to supervise the training of a final distilled architecture for real-time deployment. Extensive experiments conducted in two real-world scenes demonstrate the effectiveness of the proposed method. Specifically, we show that our pipeline enables the training of a MobileNetV3 that achieves significant mIoU gains of 3. 60% and 11. 40% on our two datasets compared to the DAFormer while only demanding 1/15 of the latter’s inference time. Code and datasets are available at https://github.com/VIS4ROB-lab/nature_uda_rt_segmentation.

EAAI Journal 2024 Journal Article

Time-segment-wise feature fusion transformer for multi-modal fault diagnosis

  • Xiaohan Zhang
  • Han Wang
  • Chenze Wang
  • Min Liu
  • Gaowei Xu

Mechanical fault diagnosis is crucial to ensure the safe operations of equipment in intelligent manufacturing systems. Recently, deep learning based fault diagnosis methods have achieved remarkable advancements with monitored data from a single sensor. However, obtaining satisfactory diagnostic results based on a single sensor is often difficult because the complementary information between different sensors is ignored. Extracting comprehensive fault features from multi-modal data is a problem that remains to be solved. To address these challenges, a time-segment-wise feature fusion Transformer (FFTR) is proposed in this paper. First, the signals from various modalities as multiple channels are normalized channel-by-channel and form a multi-modal sample. Second, a time-segment-wise feature learning network is designed to transform a multi-modal sample into several fusion features through the sequential processes of sample segmentation, segment-level feature extraction and time-aligned feature fusion. Finally, a Transformer network is employed for comprehensive multi-modal feature analysis and fault classification. In addition, a joint loss function is designed to comprehensively train the end-to-end FFTR. The comparison experiment with other baseline methods is conducted on two multi-modal datasets. The experimental results show that FFTR achieves 3. 69% and 3. 93% higher diagnostic accuracy than baseline on two datasets respectively and can address real-world problems effectively.

JBHI Journal 2023 Journal Article

Application of Robust Zero-Watermarking Scheme Based on Federated Learning for Securing the Healthcare Data

  • Baoru Han
  • Rutvij H. Jhaveri
  • Han Wang
  • Dawei Qiao
  • Jinglong Du

The privacy protection and data security problems existing in the healthcare framework based on the Internet of Medical Things (IoMT) have always attracted much attention and need to be solved urgently. In the teledermatology healthcare framework, the smartphone can acquire dermatology medical images for remote diagnosis. The dermatology medical image is vulnerable to attacks during transmission, resulting in malicious tampering or privacy data disclosure. Therefore, there is an urgent need for a watermarking scheme that doesn't tamper with the dermatology medical image and doesn't disclose the dermatology healthcare data. Federated learning is a distributed machine learning framework with privacy protection and secure encryption technology. Therefore, this paper presents a robust zero-watermarking scheme based on federated learning to solve the privacy and security issues of the teledermatology healthcare framework. This scheme trains the sparse autoencoder network by federated learning. The trained sparse autoencoder network is applied to extract image features from the dermatology medical image. Image features are undergone to two-dimensional Discrete Cosine Transform (2D-DCT) in order to select low-frequency transform coefficients for creating zero-watermarking. Experimental results show that the proposed scheme has more robustness to the conventional attack and geometric attack and achieves superior performance when compared with other zero-watermarking schemes. The proposed scheme is suitable for the specific requirements of medical images, which neither changes the important information contained in medical images nor divulges privacy data.

IJCAI Conference 2023 Conference Paper

Evaluating GPT-3 Generated Explanations for Hateful Content Moderation

  • Han Wang
  • Ming Shan Hee
  • Md Rabiul Awal
  • Kenny Tsu Wei Choo
  • Roy Ka-Wei Lee

Recent research has focused on using large language models (LLMs) to generate explanations for hate speech through fine-tuning or prompting. Despite the growing interest in this area, these generated explanations' effectiveness and potential limitations remain poorly understood. A key concern is that these explanations, generated by LLMs, may lead to erroneous judgments about the nature of flagged content by both users and content moderators. For instance, an LLM-generated explanation might inaccurately convince a content moderator that a benign piece of content is hateful. In light of this, we propose an analytical framework for examining hate speech explanations and conducted an extensive survey on evaluating such explanations. Specifically, we prompted GPT-3 to generate explanations for both hateful and non-hateful content, and a survey was conducted with 2, 400 unique respondents to evaluate the generated explanations. Our findings reveal that (1) human evaluators rated the GPT-generated explanations as high quality in terms of linguistic fluency, informativeness, persuasiveness, and logical soundness, (2) the persuasive nature of these explanations, however, varied depending on the prompting strategy employed, and (3) this persuasiveness may result in incorrect judgments about the hatefulness of the content. Our study underscores the need for caution in applying LLM-generated explanations for content moderation. Code and results are available at https: //github. com/Social-AI-Studio/GPT3-HateEval.

NeurIPS Conference 2023 Conference Paper

Improved Communication Efficiency in Federated Natural Policy Gradient via ADMM-based Gradient Updates

  • Guangchen Lan
  • Han Wang
  • James Anderson
  • Christopher Brinton
  • Vaneet Aggarwal

Federated reinforcement learning (FedRL) enables agents to collaboratively train a global policy without sharing their individual data. However, high communication overhead remains a critical bottleneck, particularly for natural policy gradient (NPG) methods, which are second-order. To address this issue, we propose the FedNPG-ADMM framework, which leverages the alternating direction method of multipliers (ADMM) to approximate global NPG directions efficiently. We theoretically demonstrate that using ADMM-based gradient updates reduces communication complexity from $\mathcal{O}({d^{2}})$ to $\mathcal{O}({d})$ at each iteration, where $d$ is the number of model parameters. Furthermore, we show that achieving an $\epsilon$-error stationary convergence requires $\mathcal{O}(\frac{1}{(1-\gamma)^{2}{\epsilon}})$ iterations for discount factor $\gamma$, demonstrating that FedNPG-ADMM maintains the same convergence rate as standard FedNPG. Through evaluation of the proposed algorithms in MuJoCo environments, we demonstrate that FedNPG-ADMM maintains the reward performance of standard FedNPG, and that its convergence rate improves when the number of federated agents increases.

NeurIPS Conference 2023 Conference Paper

Look Beneath the Surface: Exploiting Fundamental Symmetry for Sample-Efficient Offline RL

  • Peng Cheng
  • Xianyuan Zhan
  • Zhihao Wu
  • Wenjia Zhang
  • Youfang Lin
  • Shou cheng Song
  • Han Wang
  • Li Jiang

Offline reinforcement learning (RL) offers an appealing approach to real-world tasks by learning policies from pre-collected datasets without interacting with the environment. However, the performance of existing offline RL algorithms heavily depends on the scale and state-action space coverage of datasets. Real-world data collection is often expensive and uncontrollable, leading to small and narrowly covered datasets and posing significant challenges for practical deployments of offline RL. In this paper, we provide a new insight that leveraging the fundamental symmetry of system dynamics can substantially enhance offline RL performance under small datasets. Specifically, we propose a Time-reversal symmetry (T-symmetry) enforced Dynamics Model (TDM), which establishes consistency between a pair of forward and reverse latent dynamics. TDM provides both well-behaved representations for small datasets and a new reliability measure for OOD samples based on compliance with the T-symmetry. These can be readily used to construct a new offline RL algorithm (TSRL) with less conservative policy constraints and a reliable latent space data augmentation procedure. Based on extensive experiments, we find TSRL achieves great performance on small benchmark datasets with as few as 1% of the original samples, which significantly outperforms the recent offline RL algorithms in terms of data efficiency and generalizability. Code is available at: https: //github. com/pcheng2/TSRL

ICRA Conference 2023 Conference Paper

MMRDN: Consistent Representation for Multi-View Manipulation Relationship Detection in Object-Stacked Scenes

  • Han Wang
  • Jiayuan Zhang
  • Lipeng Wan 0003
  • Xingyu Chen
  • Xuguang Lan
  • Nanning Zheng 0001

Manipulation relationship detection (MRD) aims to guide the robot to grasp objects in the right order, which is important to ensure the safety and reliability of grasping in object stacked scenes. Previous works infer manipulation relationship by deep neural network trained with data collected from a predefined view, which has limitation in visual dislocation in unstructured environments. Multi-view data provide more comprehensive information in space, while a challenge of multi-view MRD is domain shift. In this paper, we propose a novel multi-view fusion framework, namely multi-view MRD network (MMRDN), which is trained by 2D and 3D multi-view data. We project the 2D data from different views into a common hidden space and fit the embeddings with a set of Von-Mises-Fisher distributions to learn the consistent representations. Besides, taking advantage of position information within the 3D data, we select a set of $K$ Maximum Vertical Neighbors (KMVN) points from the point cloud of each object pair, which encodes the relative position of these two objects. Finally, the features of multi-view 2D and 3D data are concatenated to predict the pairwise relationship of objects. Experimental results on the challenging REGRAD dataset show that MMRDN outperforms the state-of-the-art methods in multi-view MRD tasks. The results also demonstrate that our model trained by synthetic data is capable to transfer to real-world scenarios.

ICLR Conference 2023 Conference Paper

Replay Memory as An Empirical MDP: Combining Conservative Estimation with Experience Replay

  • Hongming Zhang 0003
  • Chenjun Xiao
  • Han Wang
  • Jun Jin 0001
  • Bo Xu 0002
  • Martin Müller 0003

Experience replay, which stores transitions in a replay memory for repeated use, plays an important role of improving sample efficiency in reinforcement learning. Existing techniques such as reweighted sampling, episodic learning and reverse sweep update further process the information in the replay memory to make experience replay more efficient. In this work, we further exploit the information in the replay memory by treating it as an empirical \emph{Replay Memory MDP (RM-MDP)}. By solving it with dynamic programming, we learn a conservative value estimate that \emph{only} considers transitions observed in the replay memory. Both value and policy regularizers based on this conservative estimate are developed and integrated with model-free learning algorithms. We design the metric \textit{memory density} to measure the quality of RM-MDP. Our empirical studies quantitatively find a strong correlation between performance improvement and memory density. Our method combines \emph{Conservative Estimation with Experience Replay (CEER)}, improving sample efficiency by a large margin, especially when the memory density is high. Even when the memory density is low, such a conservative estimate can still help to avoid suicidal actions and thereby improve performance.

EAAI Journal 2023 Journal Article

Thermal images-aware guided early fusion network for cross-illumination RGB-T salient object detection

  • Han Wang
  • Kechen Song
  • Liming Huang
  • Hongwei Wen
  • Yunhui Yan

RGB-T salient object detection (SOD) has been developed rapidly and achieved excellent results in recent years. However, some problems have not yet been solved. The current RGB-T datasets contain only a tiny amount of low-illumination data. The RGB-T SOD method trained based on these RGB-T datasets does not detect the salient objects in extremely low-illumination scenes very well. To improve the detection performance of low-illumination data, we can spend a lot of labor to label low-illumination data, but we tried a new idea to solve the problem by making full use of the properties of Thermal (T) images. Therefore, we propose a T-aware guided early fusion network for cross-illumination salient object detection. Specifically, in the training and testing stage, we use normal illumination data to train our network and then use low and extremely low-illumination data to verify the effectiveness of our method. In the early fusion stage, we propose a T-aware guided module (T-aware) for enhancing salient regions of RGB images at different illumination levels. Secondly, in the decoding stage, we use T images to guide the cross-modal fusion of RGB and T images. In addition, we propose a cross-modal fusion localization-remote correction module (CFL-RCM), which is used to deeply screen and correct redundant information generated by illumination variations. Comparative experiments on the VDT-2048 dataset validate the superior performance of our method on the cross-illumination RGB-T saliency detection. We also obtained favorable results on generalizability experiments with VT5000, VT1000, and VT821 datasets.

ICLR Conference 2022 Conference Paper

Multitask Prompted Training Enables Zero-Shot Task Generalization

  • Victor Sanh
  • Albert Webson
  • Colin Raffel
  • Stephen H. Bach
  • Lintang Sutawika
  • Zaid Alyafeai
  • Antoine Chaffin
  • Arnaud Stiegler

Large language models have recently been shown to attain reasonable zero-shot generalization on a diverse set of tasks (Brown et al., 2020). It has been hypothesized that this is a consequence of implicit multitask learning in language models’ pretraining (Radford et al., 2019). Can zero-shot generalization instead be directly induced by explicit multitask learning? To test this question at scale, we develop a system for easily mapping any natural language tasks into a human-readable prompted form. We convert a large set of supervised datasets, each with multiple prompts with diverse wording. These prompted datasets allow for benchmarking the ability of a model to perform completely unseen tasks specified in natural language. We fine-tune a pretrained encoder-decoder model (Raffel et al., 2020; Lester et al., 2021) on this multitask mixture covering a wide variety of tasks. The model attains strong zero-shot performance on several datasets, often outperforming models 16× its size. Further, our model attains strong performance on a subset of tasks from the BIG-Bench benchmark, outperforming models 6× its size. All trained models are available at https://github.com/bigscience-workshop/t-zero, and all prompts are available at https://github.com/bigscience-workshop/promptsource.

TMLR Journal 2022 Journal Article

No More Pesky Hyperparameters: Offline Hyperparameter Tuning for RL

  • Han Wang
  • Archit Sakhadeo
  • Adam M White
  • James M Bell
  • Vincent Liu
  • Xutong Zhao
  • Puer Liu
  • Tadashi Kozuno

The performance of reinforcement learning (RL) agents is sensitive to the choice of hyperparameters. In real-world settings like robotics or industrial control systems, however, testing different hyperparameter configurations directly on the environment can be financially prohibitive, dangerous, or time consuming. We focus on hyperparameter tuning from offline logs of data, to fully specify the hyperparameters for an RL agent that learns online in the real world. The approach is conceptually simple: we first learn a model of the environment from the offline data, which we call a calibration model, and then simulate learning in the calibration model to identify promising hyperparameters. Though such a natural idea is (likely) being used in industry, it has yet to be systematically investigated. We identify several criteria to make this strategy effective, and develop an approach that satisfies these criteria. We empirically investigate the method in a variety of settings to identify when it is effective and when it fails.

JBHI Journal 2020 Journal Article

Robust Fovea Localization Based on Symmetry Measure

  • Xiaoxin Guo
  • Han Wang
  • Xinfeng Lu
  • Xiaoying Hu
  • Songtian Che
  • Yinan Lu

Automatic fovea localization is a challenging issue. In this article, we focus on the study of fovea localization and propose a robust fovea localization method. We propose concentric circular sectional symmetry measure (CCSSM) for symmetry axis detection, and region of interest (ROI) determination, which is a global feature descriptor robust against local feature changes, to solve the lesion interference issue, i. e. , fovea visibility interference from lesions, using both structure features and morphological features. We propose the index of convexity and concavity (ICC) as the convexity-concavity measure of the surface and provide a quantitative evaluation tool for ophthalmologists to learn whether the occurrence of lesion within the ROI. We propose the weighted gradient accumulation map, which is insensitive to local intensity changes and can overcome the influence of noise and contamination, to perform refined localization. The advantages of the proposed method lies in two aspects. First, the accuracy and robustness can be achieved without typical sophisticated manner, i. e. , blood vessel segmentation and parabola fitting. Second, the lesion interference is considered in our plan of fovea localization. Our proposed symmetry-based method is innovative in the solution of fovea detection, and it is simple, practical, and controllable. Experiment results show that the proposed method can resist the interference of unbalanced illumination and lesions, and achieve high accuracy rate in five datasets. Compared to the state-of-the-art methods, high robustness and accuracy of the proposed method guarantees its reliability.

JBHI Journal 2019 Journal Article

Identifying Brain Networks at Multiple Time Scales via Deep Recurrent Neural Network

  • Yan Cui
  • Shijie Zhao
  • Han Wang
  • Li Xie
  • Yaowu Chen
  • Junwei Han
  • Lei Guo
  • Fan Zhou

For decades, task functional magnetic resonance imaging has been a powerful noninvasive tool to explore the organizational architecture of human brain function. Researchers have developed a variety of brain network analysis methods for task fMRI data, including the general linear model, independent component analysis, and sparse representation methods. However, these shallow models are limited in faithful reconstruction and modeling of the hierarchical and temporal structures of brain networks, as demonstrated in more and more studies. Recently, recurrent neural networks (RNNs) exhibit great ability of modeling hierarchical and temporal dependence features in the machine learning field, which might be suitable for task fMRI data modeling. To explore such possible advantages of RNNs for task fMRI data, we propose a novel framework of a deep recurrent neural network (DRNN) to model the functional brain networks from task fMRI data. Experimental results on the motor task fMRI data of Human Connectome Project 900 subjects release demonstrated that the proposed DRNN can not only faithfully reconstruct functional brain networks, but also identify more meaningful brain networks with multiple time scales which are overlooked by traditional shallow models. In general, this work provides an effective and powerful approach to identifying functional brain networks at multiple time scales from task fMRI data.

NeurIPS Conference 2018 Conference Paper

End-to-end Symmetry Preserving Inter-atomic Potential Energy Model for Finite and Extended Systems

  • Linfeng Zhang
  • Jiequn Han
  • Han Wang
  • Wissam Saidi
  • Roberto Car
  • Weinan E

Machine learning models are changing the paradigm of molecular modeling, which is a fundamental tool for material science, chemistry, and computational biology. Of particular interest is the inter-atomic potential energy surface (PES). Here we develop Deep Potential - Smooth Edition (DeepPot-SE), an end-to-end machine learning-based PES model, which is able to efficiently represent the PES for a wide variety of systems with the accuracy of ab initio quantum mechanics models. By construction, DeepPot-SE is extensive and continuously differentiable, scales linearly with system size, and preserves all the natural symmetries of the system. Further, we show that DeepPot-SE describes finite and extended systems including organic molecules, metals, semiconductors, and insulators with high fidelity.

ICRA Conference 2010 Conference Paper

Surface tension driven water strider robot using circular footpads

  • Onur Özcan
  • Han Wang
  • Jonathan D. Taylor
  • Metin Sitti

Water strider insects have attracted many researchers' attention with their power efficient and agile water surface locomotion. This study proposes a new water strider insect inspired robot, called STRIDE II, which uses new circular footpads for high lift, stability, payload capability, and a new elliptical leg rotation mechanism for more efficient water surface propulsion. The lift, drag and propulsion forces and the energy efficiency of this robot are modeled and experiments are conducted to verify these models. A maximum lift capacity of 53 grams is achieved with a total of 12 footpads, each 4. 2 cm in diameter for a robot weighing 21. 75 grams. For this robot, a propulsion efficiency of 22. 3% is measured. Maximum forward and turning speeds of the robot are measured as 71. 5 mm/sec and 0. 21 rad/sec, respectively. These water strider robots could be used in water surface monitoring, cleaning, and analysis in lakes, dams, rivers and sea.

v2026.09.13