Xun Kyrie Yang

Robotics · Whole-body control · Reinforcement learning

I am Xun Kyrie Yang, a visiting student in the ARG at the Oxford Robotics Institute, supervised by Prof. Ioannis Havoutis.

I completed my M.Phil. in Artificial Intelligence at The Hong Kong University of Science and Technology (Guangzhou) in October 2026, supervised by Prof. Junwei Liang. Previously, I received my B.Eng. in Automation from Harbin Institute of Technology, Shenzhen.

My research focuses on humanoid robots, whole-body control, and reinforcement learning. I am interested in combining perception, motion imitation, and learned world models to enable agile physical interaction and collaborative loco-manipulation.

I am currently working on CatchItMan, a perception-driven framework for humanoid ball catching, and MultiMimic, which explores coordinated multi-humanoid loco-manipulation. My previous work includes humanoid tennis, safe locomotion, autonomous loco-manipulation, and whole-body teleoperation.

Research in action

CatchItMan interactive simulation showing a G1 humanoid, a ball, and its predicted flight path
CatchItMan ↗
Humanoid ball catching · Try the interactive demo
ATHLETE humanoid robot holding a tennis racket during a real-world rally
ATHLETE ↗
Reactive humanoid tennis · Real-world experiments
GS-Playground simulator showcasing humanoid, quadruped, and wheeled robot embodiments
GS-Playground ↗
Photorealistic simulation · Vision-informed robot learning

Research interests

  • Perception-driven whole-body control and dynamic humanoid interaction
  • Reinforcement learning, motion imitation, and sim-to-real transfer
  • Multi-robot collaboration and world models for loco-manipulation

I also share Isaac Lab tutorials on Bilibili, with 3k subscribers and over 100k views.

Download my CV · Research demos · Email

selected publications

  1. ATHLETE: Learning Reactive Humanoid Tennis via Trajectory-Guided Motion Matching
    Zhihao Tian, Zifan Wang, Enzheng Zhu, and 4 more authors
    Under review, 2026
    Zhihao Tian and Zifan Wang contributed equally.
    arXiv HTML PDF
  2. One Motion, Diverse Strokes: A Kinematics-Aware Adaptive Event Prior for Vision-Guided Humanoid Tennis Striking
    Xun Yang and others
    2026
    Under review.
  3. GS-Playground: A High-Throughput Photorealistic Simulator for Vision-Informed Robot Learning
    Yufei Jia, Heng Zhang, Ziheng Zhang, and 39 more authors
    In Robotics: Science and Systems (RSS), 2026
    arXiv HTML PDF
  4. End-to-End Humanoid Robot Safe and Comfortable Locomotion Policy
    Zifan Wang, Xun Yang, Jianzhuang Zhao, and 5 more authors
    arXiv preprint arXiv:2508.07611, 2025
    Zifan Wang and Xun Yang contributed equally.
    arXiv HTML PDF
  5. Omni-Perception: Omnidirectional Collision Avoidance for Legged Locomotion in Dynamic Environments
    Zifan Wang, Teli Ma, Yufei Jia, and 5 more authors
    In Conference on Robot Learning (CoRL), 2025
    Oral presentation, Top 5%.
    arXiv HTML PDF