Spaces:
Running
Running
| title: README | |
| emoji: π | |
| colorFrom: red | |
| colorTo: pink | |
| sdk: static | |
| pinned: false | |
| <p align="center"> | |
| <img src="baai_humanoid_logo.png" width="720" alt="BAAI-Humanoid logo"> | |
| </p> | |
| <p align="center"> | |
| <b>Whole-body Mobile Manipulation for Humanoid Robots</b><br/> | |
| Locomotion β’ Whole-body Control β’ Contact-rich Interaction β’ Dexterous Manipulation β’ Teleoperation β’ Sim-to-Real | |
| </p> | |
| --- | |
| ## What we do | |
| We develop learning-based systems for **whole-body mobile manipulation** on humanoids β enabling robots to | |
| **move, reach, interact, and recover** robustly in complex real-world environments. | |
| ## Research themes | |
| - **Whole-body mobile manipulation**: coordinated locomotion + manipulation, task-centric control | |
| - **Contact-rich whole-body control**: balance, recovery, physical interaction, safety-aware execution | |
| - **Dexterous manipulation**: bimanual skills, multimodal policies (vision / touch / proprioception) | |
| - **Teleoperation & data**: scalable demonstration capture, retargeting, motion tracking, dataset curation | |
| - **Sim-to-real deployment**: adaptation, residual learning, reproducible training/eval/deployment toolchains | |
| ## Links | |
| - GitHub: https://github.com/BAAI-Humanoid | |