01 / 03
HumanoidNDM: Closing the Training-Deployment Gap of Motion Tracking via Humanoid Neural Dynamics Model
A real-world adaptation framework for humanoid motion tracking using a learned neural dynamics model, without motion capture.
ROBOTICS & EMBODIED INTELLIGENCE
Zhenrong
Wang.
Humanoid robots.
For the real world.
I study humanoid robots and how to make them more capable in the real world.
01 / INTRODUCTIONSCROLL TO EXPLORE ↓
01 ABOUT ME
Toward capable
humanoid intelligence.
I am currently a first-year Ph.D. student at CASIA, Galbot, and EPIC Lab (2026–present), jointly supervised by Prof. Li Yi and Prof. He Wang.
My research lies at the intersection of humanoid agility, whole-body control, and intelligent manipulation. I am interested in learning methods that connect dynamic movement and physical interaction, enabling humanoids to acquire versatile skills and adapt to the real world.
HUMANOID AGILITYWHOLE-BODY CONTROLINTELLIGENT MANIPULATIONSIM-TO-REAL
Education
- 2026–PRESENTChinese Academy of Sciences, Institute of AutomationPh.D. student
- 2022–2026Shenzhen UniversityUndergraduate · Electronic Information Engineering
02 SELECTED WORK
Research.
Exploring agile motion, physical interaction, and learning for the real world.
01 / 03
02 / 03
Track Any Motions under Any Disturbances
Humanoid motion tracking under diverse motions and disturbances.
03 / 03
End-to-End HOI Reconstruction Transformer with Graph-based Encoding
Reconstructing human–object interactions with transformer and graph-based representations.
* Equal contribution
03 CONTACT
Let’s make things
move.
Interested in robotics, motion, or working together? I’d love to hear from you.
w2271762553@gmail.comFIND ME ELSEWHEREGitHub ↗