Search: onr:"swepub:oai:DiVA.org:kth-306450" >
Meta Reinforcement ...
Meta Reinforcement Learning for Sim-to-real Domain Adaptation
-
- Arndt, Karol (author)
- Aalto Univ, Espoo, Finland.
-
- Hazara, Murtaza (author)
- Aalto Univ, Espoo, Finland.;Katholieke Univ Leuven, Dept Mech Engn, Leuven, Belgium.;Flanders Make, Robot Core Lab, Lommel, Belgium.
-
- Ghadirzadeh, Ali (author)
- KTH,Robotik, perception och lärande, RPL,Aalto Univ, Espoo, Finland
-
show more...
-
- Kyrki, Ville (author)
- Aalto Univ, Espoo, Finland.
-
show less...
-
Aalto Univ, Espoo, Finland Aalto Univ, Espoo, Finland.;Katholieke Univ Leuven, Dept Mech Engn, Leuven, Belgium.;Flanders Make, Robot Core Lab, Lommel, Belgium. (creator_code:org_t)
- IEEE, 2020
- 2020
- English.
-
In: 2020 IEEE International Conference On Robotics And Automation (ICRA). - : IEEE. ; , s. 2725-2731
- Related links:
-
https://urn.kb.se/re...
-
show more...
-
https://doi.org/10.1...
-
show less...
Abstract
Subject headings
Close
- Modern reinforcement learning methods suffer from low sample efficiency and unsafe exploration, making it infeasible to train robotic policies entirely on real hardware. In this work, we propose to address the problem of sim-to-real domain transfer by using meta learning to train a policy that can adapt to a variety of dynamic conditions, and using a task-specific trajectory generation model to provide an action space that facilitates quick exploration. We evaluate the method by performing domain adaptation in simulation and analyzing the structure of the latent space during adaptation. We then deploy this policy on a KUKA LBR 4+ robot and evaluate its performance on a task of hitting a hockey puck to a target. Our method shows more consistent and stable domain adaptation than the baseline, resulting in better overall performance.
Subject headings
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datorseende och robotik (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Vision and Robotics (hsv//eng)
Publication and Content Type
- ref (subject category)
- kon (subject category)
To the university's database