Edinburgh Research Archive

Behaviour Generation in Humanoids by Learning Potential-based Policies from Constrained Motion

Simple item page
dc.contributor.author
Howard, Matthew
en
dc.contributor.author
Klanke, Stefan
en
dc.contributor.author
Gienger, Michael
en
dc.contributor.author
Goerick, Christian
en
dc.contributor.author
Vijayakumar, Sethu
en
dc.date.accessioned
2010-08-31T13:42:09Z
dc.date.available
2010-08-31T13:42:09Z
dc.date.issued
2008-12
dc.date.updated
2010-08-31T13:42:09Z
dc.description.abstract
Movement generation that is consistent with observed or demonstrated behaviour is an efficient way to seed movement planning in complex, high-dimensionalmovement systems like humanoid robots.We present a method for learning potentialbased policies from constrained motion data. In contrast to previous approaches to direct policy learning, our method can combine observations from a variety of contexts where different constraints are in force, to learn the underlying unconstrained policy in form of its potential function. This allows us to generalise and predict behaviour where novel constraints apply. We demonstrate our approach on systems of varying complexity, including kinematic data from the ASIMO humanoid robot with 22 degrees of freedom.
en
dc.extent.pageNumbers
195-211
en
dc.identifier.doi
10.1080/11762320902789830
dc.identifier.eIssn
1754-2103
dc.identifier.issn
1176-2322
dc.identifier.uri
http://www.informaworld.com/smpp/content~db=all?content=10.1080/11762320902789830
dc.identifier.uri
http://hdl.handle.net/1842/3695
dc.language.iso
en
dc.publisher
Taylor and Francis
en
dc.subject
imitation learning
en
dc.subject
constrained motion
en
dc.subject
kinematics
en
dc.subject
dynamics
en
dc.title
Behaviour Generation in Humanoids by Learning Potential-based Policies from Constrained Motion
en
dc.type
Article
en
rps.issue
4
en
rps.title
Applied Bionics and Biomechanics
en
rps.volume
5
en

Files

Original bundle

Now showing 1 - 1 of 1
Name:
Behaviour generation in humanoids by learning potential-based policies from constrained motion.pdf
Size:
1.87 MB
Format:
Adobe Portable Document Format
Download

This item appears in the following Collection(s)

Informatics Publications