RoboCopilot: Human-in-the-loop Interactive Imitation Learning for Robot Manipulation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915190085255168 |
|---|---|
| author | Wu, Philipp Shentu, Yide Liao, Qiayuan Jin, Ding Guo, Menglong Sreenath, Koushil Lin, Xingyu Abbeel, Pieter |
| author_facet | Wu, Philipp Shentu, Yide Liao, Qiayuan Jin, Ding Guo, Menglong Sreenath, Koushil Lin, Xingyu Abbeel, Pieter |
| contents | Learning from human demonstration is an effective approach for learning complex manipulation skills. However, existing approaches heavily focus on learning from passive human demonstration data for its simplicity in data collection. Interactive human teaching has appealing theoretical and practical properties, but they are not well supported by existing human-robot interfaces. This paper proposes a novel system that enables seamless control switching between human and an autonomous policy for bi-manual manipulation tasks, enabling more efficient learning of new tasks. This is achieved through a compliant, bilateral teleoperation system. Through simulation and hardware experiments, we demonstrate the value of our system in an interactive human teaching for learning complex bi-manual manipulation skills. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_07771 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RoboCopilot: Human-in-the-loop Interactive Imitation Learning for Robot Manipulation Wu, Philipp Shentu, Yide Liao, Qiayuan Jin, Ding Guo, Menglong Sreenath, Koushil Lin, Xingyu Abbeel, Pieter Robotics Learning from human demonstration is an effective approach for learning complex manipulation skills. However, existing approaches heavily focus on learning from passive human demonstration data for its simplicity in data collection. Interactive human teaching has appealing theoretical and practical properties, but they are not well supported by existing human-robot interfaces. This paper proposes a novel system that enables seamless control switching between human and an autonomous policy for bi-manual manipulation tasks, enabling more efficient learning of new tasks. This is achieved through a compliant, bilateral teleoperation system. Through simulation and hardware experiments, we demonstrate the value of our system in an interactive human teaching for learning complex bi-manual manipulation skills. |
| title | RoboCopilot: Human-in-the-loop Interactive Imitation Learning for Robot Manipulation |
| topic | Robotics |
| url | https://arxiv.org/abs/2503.07771 |