publications

2026

1. Under Review

   

   The MI-Motion Dataset and Benchmark for 3D Multi-Person Motion Prediction

   Xiao Zhou, Xiaogang Peng, Yikai Luo, and 1 more author

   Applied Soft Computing, 2026

   Under Review

   Abs Bib

   We introduce the MI-Motion dataset, a large-scale benchmark for 3D multi-person motion prediction, along with a novel baseline that achieves state-of-the-art performance.

   @article{peng2026mimotion,
     title = {The MI-Motion Dataset and Benchmark for 3D Multi-Person Motion Prediction},
     author = {Zhou, Xiao and Peng, Xiaogang and Luo, Yikai and Wu, Zizhao},
     journal = {Applied Soft Computing},
     note = {Under Review},
     year = {2026},
   }

2. Siggraph Asia

   

   LiveDub: Real-Time Lip Sync via MambaUNet with Plug-and-Play Dental Refinement

   Xiao Zhou and others

   In SIGGRAPH Asia 2026 (Under Preparation), 2026

   Manuscript in preparation

   Abs Bib Website

   We present LiveDub, a real-time lip synchronization framework based on MambaUNet architecture with a plug-and-play dental refinement module, targeting submission to SIGGRAPH Asia 2026.

   @inproceedings{zhou2026livedub,
     title = {LiveDub: Real-Time Lip Sync via MambaUNet with Plug-and-Play Dental Refinement},
     author = {Zhou, Xiao and others},
     booktitle = {SIGGRAPH Asia 2026 (Under Preparation)},
     year = {2026},
     note = {Manuscript in preparation},
   }

2024

1. IVC

   

   Learning Weakly Supervised Audio-Visual Violence Detection in Hyperbolic Space

   Xiao Zhou, Xiaogang Peng, Hao Wen, and 4 more authors

   Image and Vision Computing, Nov 2024

   Abs DOI Bib

   We propose a novel framework that learns snippet embeddings in hyperbolic space to improve audio-visual violence detection under weak supervision, achieving state-of-the-art performance on multiple benchmarks.

   @article{zhou2024hyperVD,
     title = {Learning Weakly Supervised Audio-Visual Violence Detection in Hyperbolic Space},
     author = {Zhou, Xiao and Peng, Xiaogang and Wen, Hao and Luo, Yikai and Yu, Keyang and Yang, Ping and Wu, Zizhao},
     journal = {Image and Vision Computing},
     volume = {151},
     pages = {105286},
     year = {2024},
     month = nov,
     doi = {10.1016/j.imavis.2024.105286},
     url = {https://doi.org/10.1016/j.imavis.2024.105286},
   }