Miaowei (Michael) Wang

PhD Candidate, Engineer, Tutor, Researcher

prof_pic.jpg
Pronouns: He/Him/His
m.wang-123@sms.ed.ac.uk
University of Edinburgh

Miaowei is currently a PhD candidate at the School of Informatics, University of Edinburgh, starting in October 2023. His research focuses on controllable motion representation in computer vision and graphics. His PhD journey is fortunately advised by Prof. Amir Vaxman and Prof. Oisin Mac Aodha.

Previously, he completed his graduate studies at the Department of Electrical Engineering and Computer Science, University of Michigan, under the supervision of Prof. Jason Corso. He also collaborated with Prof. Daniel Morris at Michigan State University on research in 3D point clouds.

Besdies, he has strong industrial experience in computer vision, computer graphics, and machine learning, having worked at Tencent LightSpeed (Algorithm Intern), SenseTime (AI Researcher), ManyCore (AI Researcher), Kuaishou (Algorithm Intern), and China Telecom (Algorithm Intern), et al.

He's always open to collaborating—feel free to reach out!

Research Topics: World Models · 4D Generation · Dynamic Reconstruction

Profile image photographed at National Galleries Scotland, March 2025.

News

Feb 24, 2026 My brand new personal webpage is now live! 🎉 Check it out at wangmiaowei.github.io. Excited to share my research and projects with you!
Feb 21, 2026 Our BiMotion paper has been accepted to CVPR 2026 with strong reviewer scores of 5/6, 5/6, and 5/6. Congratulations to all coauthors! 🎉📄✨
Jan 20, 2026 Traveling to Singapore to present our AAAI 2026 papers: MotionPhysics and EvoEmpirBench. See you there! 🌏
Dec 15, 2025 Invited and sponsored by Huawei Hong Kong Research Center, attending SIGGRAPH Asia 2025 in Hong Kong. Looking forward to the conference! ⭐

Representative Publications

For a complete list of publications, please refer to Google Scholar

World Models

  1. MotionPhysics: Learnable Motion Distillation for Text-Guided Simulation
    Miaowei Wang, Jakub Zadrożny, Oisin Mac Aodha, and Amir Vaxman
    In Proceedings of the AAAI Conference on Artificial Intelligence, 2026
    arXiv Project Page Code

    TL;DR: A framework for learning controllable motion representations through distillation for physics-based simulations

4D Generation

  1. BiMotion: B-spline Motion for Text-guided Dynamic 3D Character Generation
    Miaowei Wang, Qingxuan Yan, Zhi Cao, Yayuan Li, Oisin Mac Aodha, Jason J. Corso, and Amir Vaxman
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
    arXiv Project Page Code

    TL;DR: B-spline based motion representation for generating dynamic 3D characters from text descriptions

  1. Decoupledgaussian: Object-scene decoupling for physics-based interaction
    Miaowei Wang, Yibo Zhang, Weiwei Xu, Rui Ma, Changqing Zou, and Daniel Morris
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2025
    arXiv Project Page Code

    TL;DR: Decoupling objects and scenes in Gaussian representations for realistic physics-based interactions

Dynamic Reconstruction

  1. CanFields: Consolidating Diffeomorphic Flows for Non-Rigid 4D Interpolation from Arbitrary-Length Sequences
    Miaowei Wang, Changjian Li, and Amir Vaxman
    In Proceedings of the IEEE/CVF International Conference on Computer Vision, 2025
    arXiv Project Page Code

    TL;DR: Diffeomorphic flow-based approach for non-rigid 4D surface reconstruction from arbitrary-length sequences

  1. Self-annotated 3d geometric learning for smeared points removal
    Miaowei Wang and Daniel Morris
    In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2024
    arXiv Video Code

    TL;DR: Self-supervised 3D geometric learning framework for cleaning point clouds from acquisition artifacts

Mentoring

Teaching Assistant

I have served as a Teaching Assistant for several core courses at the University of Edinburgh:

Student Mentoring

I have been fortunate to mentor and support several outstanding students in their research and career development: