Efficient online adaptation with stochastic recurrent neural networks

Daniel Tanneberg, Jan Peters, Elmar Rueckert

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in Konferenzband

OriginalspracheEnglisch
Titel2017 IEEE-RAS 17th International Conference on Humanoid Robotics, Humanoids 2017
Herausgeber (Verlag)IEEE Computer Society
Seiten198-204
Seitenumfang7
ISBN (elektronisch)9781538646786
DOIs
PublikationsstatusVeröffentlicht - 22 Dez. 2017
Veranstaltung17th IEEE-RAS International Conference on Humanoid Robotics, Humanoids 2017 - Birmingham, Großbritannien / Vereinigtes Königreich
Dauer: 15 Nov. 201717 Nov. 2017

Publikationsreihe

NameIEEE-RAS International Conference on Humanoid Robots
ISSN (Print)2164-0572
ISSN (elektronisch)2164-0580

Konferenz

Konferenz17th IEEE-RAS International Conference on Humanoid Robotics, Humanoids 2017
Land/GebietGroßbritannien / Vereinigtes Königreich
OrtBirmingham
Zeitraum15/11/1717/11/17

Bibliographische Notiz

Funding Information:
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No #713010 (GOAL-Robots) and No #640554 (SKILLS4ROBOTS). 1Intelligent Autonomous Systems, Technische Universität Darmstadt, Darmstadt, Germany, {daniel,elmar}@robot-learning.de 2Robot Learning Group, Max-Planck Institute for Intelligent Systems, Tübingen, Germany, mail@jan-peters.net Fig. 1: Experimental setup. Picture of the used setup for online planning and learning, showing the KUKA LWR arm and the environment with two obstacles. The crosses depict the via points that need to be reached successively. One obstacle was pre-trained with a realistic dynamic simulation of the robot and the second is learned additionally online on the real system.

Publisher Copyright:
© 2017 IEEE.

Dieses zitieren