Junxuan Bai

Short Bio

I am a lecturer at Institute of Artificial Intelligence in Sports (IAIS), Capital University of Physical Education and Sports (CUPES), where I bridge computer graphics, VR/AR, and sports science. My research focuses on deep-learning-driven human motion modeling and its applications in athletic training, skill acquisition, and immersive sports systems.

Previously, I was a researcher at China Mobile Research Institute and obtained my Ph.D. from State Key Laboratory of Virtual Reality Technology and Systems at Beihang University.

Scholarly Service

Reviewer: IEEE VR, ISMAR, VRST, CASA, TVC, CAVW, IEEE SPL, AAAI, ECAI

Research Highlight

🧠

AI for Human Motion

Deep learning for human pose estimation, motion prediction, motion synthesis, and motion analysis.

Motion Papers →

🏀

Immersive Training

VR/AR systems for real-time training and skill learning.

Training Projects →

🎮

Character & Animation

Skinning, NeRF avatars, and deformable models—transferring CG techniques from virtual surgery to sports simulation.

CG Papers →

News

[2026-03] Our papers, "BounceCoach: Real-Time Basketball Dribble Training with On-Site Large Language Model Guidance", is accepted by 2026 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW). The information can be seen at this link.

[2025-06] Our papers, "LiteNeRFAvatar: A lightweight NeRF with local feature learning for dynamic human avatar", is accepted by Pattern Recognition. The information can be seen at this link and the demo video can be found at this link.

[2025-05] Our papers, "Motion In-Betweening via Recursive Keyframe Prediction", is published on Computer Animation and Virtual Worlds. The information can be seen at this link.

[2024-11] Our papers, "Motion Editing for Quadruped Characters via Latent Frequency Embedding", is published on IEEE Transactions on Visualization and Computer Graphics. The information can be seen at this link.

[2024-05] Our papers, "Free editing of Shape and Texture with Deformable Net for 3D Caricature Generation", is published on The Visual Computer. The information can be seen at this link.