xbpeng/DeepMimic
A physics-based character animation system that uses deep reinforcement learning to train simulated humanoids to imitate mocap motion skills.
Not currently ranked — collecting fresh signals.
star history
This research project implements deep RL algorithms for motion imitation, enabling simulated characters to learn and perform complex physical skills from motion capture data. The system supports training through adversarial motion priors and multiple imitation techniques. It is now deprecated in favor of the faster MimicKit implementation.
Frequently asked
- What is xbpeng/DeepMimic?
- A physics-based character animation system that uses deep reinforcement learning to train simulated humanoids to imitate mocap motion skills.
- Is DeepMimic open source?
- Yes — xbpeng/DeepMimic is open source, released under the MIT license.
- What language is DeepMimic written in?
- xbpeng/DeepMimic is primarily written in C++.
- How popular is DeepMimic?
- xbpeng/DeepMimic has 3.1k stars on GitHub.
- Where can I find DeepMimic?
- xbpeng/DeepMimic is on GitHub at https://github.com/xbpeng/DeepMimic.