Exciting Action: Investigating Efficient Exploration for Learning Musculoskeletal Humanoid Locomotion
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914872746311680 |
|---|---|
| author | Geiß, Henri-Jacques Al-Hafez, Firas Seyfarth, Andre Peters, Jan Tateo, Davide |
| author_facet | Geiß, Henri-Jacques Al-Hafez, Firas Seyfarth, Andre Peters, Jan Tateo, Davide |
| contents | Learning a locomotion controller for a musculoskeletal system is challenging due to over-actuation and high-dimensional action space. While many reinforcement learning methods attempt to address this issue, they often struggle to learn human-like gaits because of the complexity involved in engineering an effective reward function. In this paper, we demonstrate that adversarial imitation learning can address this issue by analyzing key problems and providing solutions using both current literature and novel techniques. We validate our methodology by learning walking and running gaits on a simulated humanoid model with 16 degrees of freedom and 92 Muscle-Tendon Units, achieving natural-looking gaits with only a few demonstrations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_11658 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Exciting Action: Investigating Efficient Exploration for Learning Musculoskeletal Humanoid Locomotion Geiß, Henri-Jacques Al-Hafez, Firas Seyfarth, Andre Peters, Jan Tateo, Davide Robotics Machine Learning Learning a locomotion controller for a musculoskeletal system is challenging due to over-actuation and high-dimensional action space. While many reinforcement learning methods attempt to address this issue, they often struggle to learn human-like gaits because of the complexity involved in engineering an effective reward function. In this paper, we demonstrate that adversarial imitation learning can address this issue by analyzing key problems and providing solutions using both current literature and novel techniques. We validate our methodology by learning walking and running gaits on a simulated humanoid model with 16 degrees of freedom and 92 Muscle-Tendon Units, achieving natural-looking gaits with only a few demonstrations. |
| title | Exciting Action: Investigating Efficient Exploration for Learning Musculoskeletal Humanoid Locomotion |
| topic | Robotics Machine Learning |
| url | https://arxiv.org/abs/2407.11658 |