Whole-Body Mobile Manipulation using Offline Reinforcement Learning on Sub-optimal Controllers
Mobile Manipulation (MoMa) of articulated objects, such as opening doors, drawers, and cupboards, demands simultaneous, whole-body coordination between a robot's base and arms. Classical whole-body controllers (WBCs) can solve such problems via hierarchical optimization, but require extensive hand-t...
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Artículo |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2604.12509 http://arxiv.org/abs/2604.12509 |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|