Shaoli Huang

myphoto.jpg

I am currently a senior researcher at Tencent AI Lab. I obtained my Ph.D. from the University of Technology Sydney under the supervision of Professor Tao Dacheng and have served as a researcher fellow at the University of Sydney. My research interests lean towards practical applications with a strong emphasis on digital human-related technologies. These include video-based human motion capture, motion generation, and motion retargeting, among others. I am passionate about exploring and developing cutting-edge solutions to drive advancements in these fields.

Project 1: Audio-driven Full-body Motion Generation
This project focuses on generating natural, stable, and rhythmically synchronized full-body motions driven by speech input.The system aims to provide a robust solution for applications in virtual reality, digital entertainment, and human-computer interaction, ensuring high fidelity and realism in audio-driven motion generation. Find more demo videos in hologest.github.io.
Project 2: Visual-based Motion Capture
This project aims to extract high-quality human motion data directly from video input. The extracted motion data can be used for video-driven 2D/3D digital human animation and the creation of large-scale 3D human motion datasets, supporting applications in virtual reality, gaming, and AI-driven character simulation.

news

Dec 10, 2024 One paper accepted by AAAI 2025! :sparkles:
Nov 05, 2024 One paper accepted by 3DV 2025! :sparkles:
Oct 10, 2024 One spotlight paper accepted by NIPS 2024! :sparkles:
Oct 05, 2024 Three papers accepted by WACV 2025! :sparkles:
Jul 01, 2024 Three papers (one Oral) accepted by ECCV 2024! :sparkles:
Apr 30, 2024 One paper accepted by ACM SIGGRAPH 2024. Featured in Technical Papers Video Trailer! :sparkles:
Feb 27, 2024 One spotlight paper accepted by CVPR 2024! :sparkles:
Jan 16, 2024 One paper accepted by ICLR 2024! :sparkles:

selected publications

  1. ECCV
    crossdiff.gif
    Realistic Human Motion Generation with Cross-Diffusion Models
    Zeping Ren, Shaoli Huang†, and Xiu Li†
    2024
    arXiv Code Website
  2. ECCV
    signavatar.gif
    SignAvatars: A Large-scale 3D Sign Language Holistic Motion Dataset and Benchmark
    Zhengdi Yu, Shaoli Huang†, Yongkang Cheng, and Togal Bridal
    2024
    arXiv Code Website
  3. SIGGRAPH
    camdm.gif
    Taming Diffusion Probabilistic Models for Character Control
    Rui Chen*, Mingyi Shi*, Shaoli Huang, Ping Tan, Taku Komura, and Xuelin Chen†
    2024
    Video Code Website
  4. AAAI
    hutu.gif
    HuTuMotion: Human-Tuned Navigation of Latent Motion Diffusion Models with Minimal Feedback
    Gaoge Han, Shaoli Huang†, Mingming Gong, and Jinglei Tang†
    2024
    Code Website
  5. ICLR
    tapmo.gif
    TapMo: Shape-aware Motion Generation of Skeleton-free Characters
    Jiaxu Zhang*, Shaoli Huang*, Zhigang Tu†, Xin Chen, Xiaohang Zhan, YU Gang, and Ying Shan
    In , 2024
    Code Website
  6. CVPR
    progmogen.gif
    Programmable Motion Generation for Open-Set Motion Control Tasks
    Hanchao Liu, Xiaohang Zhan†, Shaoli Huang, Tai-Jiang Mu†, and Ying Shan
    2024
    Code Website
  7. ICCV
    livelyspeaker.gif
    Livelyspeaker: Towards semantic-aware co-speech gesture generation
    Yihao Zhi*, Xiaodong Cun*, Xuelin Chen, Xi Shen, Wen Guo, Shaoli Huang, and Shenghua Gao†
    2023
    Code Website
  8. CVPR
    acr.gif
    ACR: Attention collaboration-based regressor for arbitrary two-hand reconstruction
    Zhengdi Yu, Shaoli Huang†, Chen Fang, Toby P Breckon, and Jue Wang
    In , 2023
    Code Website
  9. CVPR
    anchor.gif
    Learning anchor transformations for 3d garment animation
    Fang Zhao, Zekun Li, Shaoli Huang†, Junwu Weng, Tianfei Zhou, Guo-Sen Xie, Jue Wang, and Ying Shan
    In , 2023
    Website
  10. CVPR
    r2et.gif
    Skinned motion retargeting with residual perception of motion semantics & geometry
    Jiaxu Zhang, Junwu Weng, Di Kang, Fang Zhao, Shaoli Huang, Xuefei Zhe, Linchao Bao, Ying Shan, Jue Wang, and Zhigang Tu
    2023
    Code Website