Overview

Humanoid motion tracking studies how robots can reproduce human movements and follow motion commands despite differences in body shape, dynamics, and environment. Common approaches retarget motion-capture or video data to robot bodies, then train physics-based controllers with reinforcement learning or imitation; newer systems also use Transformers and generative motion models to broaden the range of skills. Research increasingly tests real-robot transfer, terrain and object interaction, teleoperation, recovery, and adaptation across robot embodiments. A central challenge is balancing faithful, expressive motion with physical feasibility, safety, and robustness, while benchmarks that measure pose error alone may miss unstable contacts or poor task execution.

Latest papers

91 papers23 papers in the last four weeks, up 35% on the four weeks before. 0.2% of all new papers.

Jun 15, 2025

KungfuBot: Physics-Based Humanoid Whole-Body Control for Learning Highly-Dynamic Skills

Humanoid robots are promising to acquire various skills by imitating human behaviors. However, existing algorithms are only capable of tracking smooth, low-speed human motions, even with delicate reward and curriculum design. This paper presents a physics-based humanoid control framework, aiming to master highly-dynamic human behaviors such as Kungfu and dancing through multi-steps motion processing and adaptive motion tracking. For motion processing, we design a pipeline to extract, filter out, correct, and retarget motions, while ensuring compliance with physical constraints to the maximum extent. For motion imitation, we formulate a bi-level optimization problem to dynamically adjust the tracking accuracy tolerance based on the current tracking error, creating an adaptive curriculum mechanism. We further construct an asymmetric actor-critic framework for policy training. In experiments, we train whole-body control policies to imitate a set of highly-dynamic motions. Our method achieves significantly lower tracking errors than existing approaches and is successfully deployed on the Unitree G1 robot, demonstrating stable and expressive behaviors. The project page is https://kungfubot.github.io.