I received my Ph.D. from Wuhan University of Technology in 2024, and I was extremely fortunate to be supervised by Prof. Luo Zhong, one of the most dedicated teachers I have ever met. My PhD research interests lie in action recognition under complex environments, video content analysis, and computing science. My ultimate aim is to use machine intelligence to help understand human action, facilitating daily life and health.

I am currently delving into spike vision at Peking University, under the supervision of Prof. Tiejun Huang, a leading expert in spike vision.

🤜🤛 Group

I maintain close collaborations with Prof. Zhong Xian’s team XIAN Group at Wuhan University of Technology and AIM Lab of Prof. Wang Zheng from Wuhan University. If you are seeking any form of academic cooperation, please feel free to email at liuwx66@pku.edu.cn.

🔥 Action-related News

• 2025.06:  🎉🎉 One paper was accepted by TITS.
• 2025.06:  🎉🎉 Three co-authored papers were accepted by ACM MM 2025.
• 2025.03:  🎉🎉 One co-authored paper was accepted by CVPR 2025.
• 2024.12:  🎉🎉 One paper was accepted by Journal of Image and Graphics, 中国图象图形学报.
• 2024.08:  🎉🎉 One paper was accepted by PR.
• 2024.07:  🎉🎉 One paper was accepted by ACM MM 2024 International Workshop on Human-centric Multimedia Analysis.
• 2024.05:  🎉🎉 One paper was accepted by The Visual Computer.
• 2024.07:  🎉🎉 One co-authored paper (时空语义驱动的渐进多视角行为去偏置研究) was accepted by Computer Engineering, 计算机工程.
• 2024.07:  🎉🎉 One co-authored paper was accepted by ACM MM 2024.
• 2024.03:  🎉🎉 One co-authored paper was accepted by CVPR 2024.
• 2023.10:  🎉🎉 One co-authored paper was accepted by ICIP.
• 2023.07:  🎉🎉 One co-authored paper was accepted by ACM MM 2023.
• 2023.06:  🎉🎉 Two co-authored papers were accepted by ICASSP 2023.
• 2023.05:  🎉🎉 One paper was accepted by TIP.
• 2022.12:  🎉🎉 One paper was accepted by SPL.
• 2022.06:  🎉🎉 One co-authored paper was accepted by ICASSP 2022.
• 2021.11:  🎉🎉 Two co-authored papers were accepted by PRICAI 2021.

🔥 Spike-related News

• 2025.05:  🎉🎉 One co-authored paper was accepted by ICML 2025.
• 2025.05:  🎉🎉 One co-authored paper was accepted by Neurocomputing.
• 2025.04:  🎉🎉 One paper was accepted by IJCAI 2025.
• 2024.07:  🎉🎉 One co-authored paper was accepted by ACM MM 2024.
• 2024.04:  🎉🎉 One co-authored paper was accepted by TCDS.

🔥 Other News

• 2025.06:  🎉🎉 One co-authored papers were accepted by ACM MM 2025
• 2024.12:  🎉🎉 One co-authored paper was accepted by CVIU (CCF B). Congratulations to YES!
• 2024.10:  🎉🎉 One co-authored paper was accepted by AAAI2025.
• 2024.07:  🎉🎉 One co-authored paper was accepted by ACM MM 2024 Workshop on Multi-modal Misinformation Governance in the Era of Foundation Models.
• 2023.11:  🎉🎉 One co-authored paper was accepted by PR.
• 2023.07:  🎉🎉 One co-authored paper was accepted by ACM MM 2023.
• 2023.06:  🎉🎉 Two co-authored papers were accepted by ICASSP 2023.
• 2022.07:  🎉🎉 One co-authored paper was accepted by ICME 2022.
• 2022.06:  🎉🎉 One co-authored paper was accepted by TIP.
• 2021.11:  🎉🎉 One co-authored paper was accepted by MMM.
• 2021.07:  🎉🎉 One co-authored paper was accepted by ACM MM 2021.
• 2020.06:  🎉🎉 One co-authored paper was accepted by ICMR 2020.

📝 Publications

IJCAI 2025

SOTA: Spike-Navigated Optimal TrAnsport Saliency Region Detection in Composite-bias Videos

Wenxuan Liu, Yao Deng, Kang Chen, Xian Zhong, Zhaofei Yu, Tiejun Huang

• Existing saliency detection methods struggle in real-world scenarios due to motion blur and occlusions. In contrast, spike cameras, with their high temporal resolution, significantly enhance visual saliency maps. However, the composite noise inherent to spike camera imaging introduces discontinuities in saliency detection. Low-quality samples further distort model predictions, leading to saliency bias. To address these challenges, we propose Spike-navigated Optimal TrAnsport Saliency Region Detection (SOTA), a framework that leverages the strengths of spike cameras while mitigating biases in both spatial and temporal dimensions. Our method introduces Spike-based Micro-debias (SM) to capture subtle frame-to-frame variations and preserve critical details, even under minimal scene or lighting changes. Additionally, Spike-based Global-debias (SG) refines predictions by reducing inconsistencies across diverse conditions. Extensive experiments on real and synthetic datasets demonstrate that SOTA outperforms existing methods by eliminating composite noise bias.

Thanks: I would like to express my sincere gratitude to Xiyu Han for the valuable assistance in preparing the figures for my paper.

PR

Dynamic and static mutual fitting for action recognition

Wenxuan Liu, Xuemei Jia, Xian Zhong, Kui Jiang, Xiaohan Yu, Mang Ye

PDF

• We propose Dynamic Temporal-aware Erasing (DTE) to excavate the fine-grained representation based on actor trajectory without destroying temporal information.

TIP

Dual-Recommendation Disentanglement Network for View Fuzz in Action Recognition

Wenxuan Liu, Xian Zhong, Zhuo Zhou, Kui Jiang, Zheng Wang, and Chia-Wen Lin

PDF CODE

• This paper perceives a ubiquitous view fuzz issue in multi-view action recognition tasks, which undermines the action feature distillation by hiding essential action regions.

ACM MM HCMA

Pixel-REfocused Navigated Tri-margin for Semi-supervised Action Detection

Wenxuan Liu, Shilei Zhao, Xiyu Han, Aoyu Yi, Kui Jiang, Zheng Wang, Xian Zhong (Best Student Runner-up)

PDF

• This paper identifies a novel issue, termed pixel activation-uncertainty, in semi-supervised action detection, which highlights the difficulty in distinguishing between action and background boundaries due to active motion.

ACM MM 2024

Predicting the Unseen: A Novel Dataset for Hidden Intention Localization in Pre-abnormal Analysis

ZeHao Qi, Ruixu Zhang, Xinyi Hu, Wenxuan Liu, Zheng Wang

PDF Dataset

• The abnormal intention is dissected into four coarse-grained dimensions, under which eleven detailed perspectives are distinguished.

• Fragrant: frequency-auxiliary guided relational attention network for low-light action recognition, Wenxuan Liu, Xuemei Jia, Yihao Ju, Yakun Ju, Kui Jiang, Shifeng Wu, Luo Zhong, Xian Zhong, The Visual Computer, 2024.
• Actor-aware alignment network for action recognition, Wenxuan Liu, Xian Zhong, Xuemei Jia, Kui Jiang, Chia-Wen Lin, IEEE Signal Processing Letters, 2022.
• YES: You should Examine Suspect Cues for Low-Light Object Detection, Shu Ye, Wenxin Huang, Wenxuan Liu, Liang Chen, Xiao Wang, Xian Zhong, CVIU
• Pioneering Explainable Video Fact-Checking with a New Dataset and Multi-role Multimodal Model Approach, Kaipeng Niu, Danni xu, Bingjian Yang, Wenxuan Liu, Zheng Wang, AAAI 2025
• Towards Low-latency Event-based Visual Recognition with Hybrid Step-wise Distillation Spiking Neural Networks, Xian Zhong, Shengwang Hu, Wenxuan Liu*, Wenxin Huang, Jianhao Ding, Zhaofei Yu, Tiejun Huang, ACM MM 2024
• Predicting the Unseen: A Novel Dataset for Hidden Intention Localization in Pre-abnormal Analysis, Zehao Qi, Ruixu Zhang, Xinyi Hu, Wenxuan Liu, Zheng Wang, ACM MM 2024
• Assess and Guide: Multi-modal Fake News Detection via Decision Uncertainty, Jie Wu, Danni Xu, Wenxuan Liu, Joey Zhou, Yew Ong, Siyuan Hu, Hongyuan Zhu, Zheng Wang, ACM MM 2024 Workshop Multi-modal Misinformation
• Bi-Causal: Group Activity Recognition via Bidirectional Causality, Youliang Zhang, Wenxuan Liu, Danni Xu, Zhuo Zhou, Zheng Wang, CVPR 2024
• ICLR: Instance Credibility-Based Label Refinement for label noisy person re-identification, Xian Zhong, Xiyu Han, Xuemei Jia, Wenxin Huang, Wenxuan Liu, Shuaipeng Su, Xiaohan Yu, Mang Ye, PR 2024
• Converting Artificial Neural Networks to Ultra-Low-Latency Spiking Neural Networks for Action Recognition, Hong You, Xian Zhong, Wenxuan Liu, Qi Wei, Wenxin Huang, Zhaofei Yu, Tiejun Huang, TCDS 2024
• Uncovering the Unseen: Discover Hidden Intentions by Micro-Behavior Graph Reasoning, Zhuo Zhou, Wenxuan Liu*, Danni Xu, Zheng Wang, Jian Zhao, ACM MM 2023
• Dawn: Direction-aware attention wavelet network for image deraining, Kui Jiang, Wenxuan Liu, Zheng Wang, Xian Zhong, Junjun Jiang, Chia-Wen Lin, ACM MM 2023
• Background-weakening consistency regularization for semi-supervised video action detection, Xian Zhong, Aoyu Yi, Wenxuan Liu*, Wenxin Huang, Chengming Zou, Zheng Wang, ICASSP 2023
• Neighborhood information-based label refinement for person re-identification with label noise, Xian Zhong, Shuaipeng Su, Wenxuan Liu*, Xuemei Jia, Wenxin Huang, Mengdie Wang, ICASSP 2023
• Implicit Attention-Based Cross-Modal Collaborative Learning for Action Recognition, Jianghao Zhang, Xian Zhong, Wenxuan Liu*, Kui Jiang, Zhengwei Yang, Zheng Wang, ICIP 2023

🎖 Honors and Awards

• 2024.10 ACM MM 2024 the 5-th International Workshop on Human-centric Multimedia Analysis, Best Student Paper Runner-Up.
• 2024.07 Outstanding Graduate Student of Wuhan University of Technology.
• 2023.07 First Prize in the ICME ’23 Grand Challenge “Seeing Through the Rain (STRAIN): Vision Task Challenges in Real-world Rain Scenes”, Track 3.
• 2022.08 Third Prize in the 11th China Software Cup Student Software Design Competition

📖 Educations

• 2024.06 - now, Postdoc, Peking University, China.
• 2019.09 - 2024.06, Ph D, Wuhan University of Technology, China.
• 2017.09 - 2019.06, Master, Wuhan University of Technology, China.
• 2012.09 - 2016.06, Bachelor, Xi’an Polytechnic of University, China.