chengxuxin/expressive-humanoid
A robotics research project developing expressive whole-body control for humanoid robots using imitation and reinforcement learning.
Not currently ranked — collecting fresh signals.
star history
This project from RSS 2024 develops whole-body control policies for humanoid robots that can express human-like motion. It combines imitation learning from motion capture data with reinforcement learning fine-tuning in Isaac Gym simulation, followed by sim-to-real transfer to physical hardware. The system uses PyTorch-based deep RL (rsl_rl, legged_gym) to train expressive motor skills.
Frequently asked
- What is chengxuxin/expressive-humanoid?
- A robotics research project developing expressive whole-body control for humanoid robots using imitation and reinforcement learning.
- Is expressive-humanoid open source?
- Yes — chengxuxin/expressive-humanoid is an open-source project tracked on heatdrop.
- What language is expressive-humanoid written in?
- chengxuxin/expressive-humanoid is primarily written in Python.
- How popular is expressive-humanoid?
- chengxuxin/expressive-humanoid has 494 stars on GitHub.
- Where can I find expressive-humanoid?
- chengxuxin/expressive-humanoid is on GitHub at https://github.com/chengxuxin/expressive-humanoid.