Sökning: onr:"swepub:oai:DiVA.org:kth-306450" > Meta Reinforcement ...
Fältnamn | Indikatorer | Metadata |
---|---|---|
000 | 02605naa a2200325 4500 | |
001 | oai:DiVA.org:kth-306450 | |
003 | SwePub | |
008 | 211217s2020 | |||||||||||000 ||eng| | |
024 | 7 | a https://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-3064502 URI |
024 | 7 | a https://doi.org/10.1109/ICRA40945.2020.91965402 DOI |
040 | a (SwePub)kth | |
041 | a engb eng | |
042 | 9 SwePub | |
072 | 7 | a ref2 swepub-contenttype |
072 | 7 | a kon2 swepub-publicationtype |
100 | 1 | a Arndt, Karolu Aalto Univ, Espoo, Finland.4 aut |
245 | 1 0 | a Meta Reinforcement Learning for Sim-to-real Domain Adaptation |
264 | 1 | b IEEE,c 2020 |
338 | a print2 rdacarrier | |
500 | a QC 20211217conference ISBN 978-1-7281-7395-5 | |
520 | a 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. | |
650 | 7 | a NATURVETENSKAPx Data- och informationsvetenskapx Datorseende och robotik0 (SwePub)102072 hsv//swe |
650 | 7 | a NATURAL SCIENCESx Computer and Information Sciencesx Computer Vision and Robotics0 (SwePub)102072 hsv//eng |
700 | 1 | a Hazara, Murtazau Aalto Univ, Espoo, Finland.;Katholieke Univ Leuven, Dept Mech Engn, Leuven, Belgium.;Flanders Make, Robot Core Lab, Lommel, Belgium.4 aut |
700 | 1 | a Ghadirzadeh, Aliu KTH,Robotik, perception och lärande, RPL,Aalto Univ, Espoo, Finland4 aut0 (Swepub:kth)u1hbw8ng |
700 | 1 | a Kyrki, Villeu Aalto Univ, Espoo, Finland.4 aut |
710 | 2 | a Aalto Univ, Espoo, Finland.b Aalto Univ, Espoo, Finland.;Katholieke Univ Leuven, Dept Mech Engn, Leuven, Belgium.;Flanders Make, Robot Core Lab, Lommel, Belgium.4 org |
773 | 0 | t 2020 IEEE International Conference On Robotics And Automation (ICRA)d : IEEEg , s. 2725-2731q <2725-2731 |
856 | 4 8 | u https://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-306450 |
856 | 4 8 | u https://doi.org/10.1109/ICRA40945.2020.9196540 |
Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.
Kopiera och spara länken för att återkomma till aktuell vy