Dynamic movement primitives in latent space of time-dependent variational autoencoders

Dynamic movement primitives (DMPs) are powerful for the generalization of movements from demonstration. However, high dimensional movements, as they are found in robotics, make finding efficient DMP representations difficult. Typically, they are either used in configuration or Cartesian space, but b...

Full description

Saved in:
Bibliographic Details
Published in:IEEE-RAS International Conference on Humanoid Robots (Print) pp. 629 - 636
Main Authors: Nutan Chen, Karl, Maximilian, van der Smagt, Patrick
Format: Conference Proceeding
Language:English
Published: IEEE 01.11.2016
Subjects:
Decoding
Gaussian distribution
Neural networks
Probabilistic logic
Switches
Training
Trajectory
ISSN:2164-0580
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Dynamic movement primitives (DMPs) are powerful for the generalization of movements from demonstration. However, high dimensional movements, as they are found in robotics, make finding efficient DMP representations difficult. Typically, they are either used in configuration or Cartesian space, but both approaches do not generalize well. Additionally, limiting DMPs to single demonstrations restricts their generalization capabilities. In this paper, we explore a method that embeds DMPs into the latent space of a time-dependent variational autoencoder framework. Our method enables the representation of high-dimensional movements in a low-dimensional latent space. Experimental results show that our framework has excellent generalization in the latent space, e.g., switching between movements or changing goals. Also, it generates optimal movements when reproducing the movements.
AbstractList Dynamic movement primitives (DMPs) are powerful for the generalization of movements from demonstration. However, high dimensional movements, as they are found in robotics, make finding efficient DMP representations difficult. Typically, they are either used in configuration or Cartesian space, but both approaches do not generalize well. Additionally, limiting DMPs to single demonstrations restricts their generalization capabilities. In this paper, we explore a method that embeds DMPs into the latent space of a time-dependent variational autoencoder framework. Our method enables the representation of high-dimensional movements in a low-dimensional latent space. Experimental results show that our framework has excellent generalization in the latent space, e.g., switching between movements or changing goals. Also, it generates optimal movements when reproducing the movements.
Author Nutan Chen
Karl, Maximilian
van der Smagt, Patrick
Author_xml – sequence: 1
  surname: Nutan Chen
  fullname: Nutan Chen
  email: nutan.chen@gmail.com
  organization: Fac. for Inf., Tech. Univ. Munchen, Munich, Germany
– sequence: 2
  givenname: Maximilian
  surname: Karl
  fullname: Karl, Maximilian
  email: karlma@in.tum.de
  organization: Fac. for Inf., Tech. Univ. Munchen, Munich, Germany
– sequence: 3
  givenname: Patrick
  surname: van der Smagt
  fullname: van der Smagt, Patrick
  organization: Fac. for Inf., Tech. Univ. Munchen, Munich, Germany
BookMark eNotkNFKwzAUhqMoOOeeQJC8QOdJ0zTp5djUDaa70Hk7TpsTiKzpaGJhby_FXX3wXXz8_PfsJnSBGHsSMBcCquf1_n3xsdusPuc5iHKuDUhZwBWbVdoIBRUUWhh1zSa5KIsMlIE7NovxBwCkMKbKywn7Xp0Dtr7hbTdQSyHxU-9bn_xAkfvAj5hGGU_YEO8cT76lzNKJgh39gL3H5LuAR46_qaPQdJb6-MBuHR4jzS6csv3ry9dynW13b5vlYpt5oVXKSl2Ay7UTJIsaoZaEpqZaggXSxuayVCgF2cYhaWcJnaugASIFhUVVyyl7_O96IjqM07E_Hy5PyD8cLlgL
ContentType Conference Proceeding
DBID 6IE
6IL
CBEJK
RIE
RIL
DOI 10.1109/HUMANOIDS.2016.7803340
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Xplore POP ALL
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP All) 1998-Present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISBN 9781509047185
1509047182
EISSN 2164-0580
EndPage 636
ExternalDocumentID 7803340
Genre orig-research
GroupedDBID 6IE
6IF
6IH
6IK
6IL
6IN
AAJGR
AAWTH
ABLEC
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
IPLJI
OCL
RIE
RIL
ID FETCH-LOGICAL-i175t-6740f27f1e34ba0b3ea8beb30d0e78d2365a31edcfae7fdeaff90c0ee504da5b3
IEDL.DBID RIE
ISICitedReferencesCount 47
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000403009300095&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Aug 27 02:00:39 EDT 2025
IsPeerReviewed false
IsScholarly false
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i175t-6740f27f1e34ba0b3ea8beb30d0e78d2365a31edcfae7fdeaff90c0ee504da5b3
PageCount 8
ParticipantIDs ieee_primary_7803340
PublicationCentury 2000
PublicationDate 2016-Nov.
PublicationDateYYYYMMDD 2016-11-01
PublicationDate_xml – month: 11
  year: 2016
  text: 2016-Nov.
PublicationDecade 2010
PublicationTitle IEEE-RAS International Conference on Humanoid Robots (Print)
PublicationTitleAbbrev HUMANOIDS
PublicationYear 2016
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0003188926
Score 1.8446746
Snippet Dynamic movement primitives (DMPs) are powerful for the generalization of movements from demonstration. However, high dimensional movements, as they are found...
SourceID ieee
SourceType Publisher
StartPage 629
SubjectTerms Decoding
Gaussian distribution
Neural networks
Probabilistic logic
Switches
Training
Trajectory
Title Dynamic movement primitives in latent space of time-dependent variational autoencoders
URI https://ieeexplore.ieee.org/document/7803340
WOSCitedRecordID wos000403009300095&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PS8MwFA5zeNCLPzbxNzl4NFvWtE1yFOeYoHOgG7uNpHmBgbZja_f3m7R1KnjxVgJt4SV5ee_lfd-H0A21PeBByEkEVpBQB5zomFtiGRhKraKmJPWZPvHRSMxmctxAt1ssDACUzWfQ8Y_lXb7JksKXyrpcUMZCl6DvcB5XWK1tPcWtTSGDuAYB96jsDifPd6OXx_6rb-CKO_XLv1RUykNkcPC_3x-i9jcaD4-358wRakB6jPZ_EAm20LRfCcvjj6zk_87x0st1eVe2xosUv7uI0g067-G-lVnsFeXJl_5tjjcuYa6LglgVeebJLX2DcxtNBg9v90NSKyaQhQsDchLzkNqAO_uzUCuqGSihXbpMDQUuTMDiSLEemMQq4NaAslbShAJENDQq0uwENdMshVOEBUjttSUNuIDD7Xulo0CCS06kv2m1yRlqeQvNlxUpxrw2zvnfwxdoz09CBeK7RM18VcAV2k02-WK9ui5n8hMT_KJ2
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PT8IwFG6ImqgXf4Dxtz14tFDWbt2ORiQQAUkEwo2062tCohuBwd9vu03UxIu3pcmWpa99fe_1fd-H0D01TRAeF8QHExKuPEFUIAwxDDSlRlKdk_pMemIwCKfTaFhBD1ssDADkzWdQd4_5Xb5O47UrlTVESBnjNkHf9Tn3aIHW2lZU7OoMIy8oYcBNGjU64_7j4LXbenMtXEG9fP2Xjkp-jLSP_vcDx6j2jcfDw-1Jc4IqkJyiwx9UglU0aRXS8vgjzRnAM7xwgl3Oma3wPMHvNqa0g9Z_2G-lBjtNefKlgJvhjU2Zy7IglussdfSWrsW5hsbt59FTh5SaCWRuA4GMBIJT4wlrAcaVpIqBDJVNmKmmIELtscCXrAk6NhKE0SCNiWhMAXzKtfQVO0M7SZrAOcIhRMqpS2qwIYfd-VL5XgQ2PYncXauJL1DVzdBsUdBizMrJufx7-A7td0b93qzXHbxcoQNnkALSd412suUabtBevMnmq-VtbtVPswKlvQ
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=IEEE-RAS+International+Conference+on+Humanoid+Robots+%28Print%29&rft.atitle=Dynamic+movement+primitives+in+latent+space+of+time-dependent+variational+autoencoders&rft.au=Nutan+Chen&rft.au=Karl%2C+Maximilian&rft.au=van+der+Smagt%2C+Patrick&rft.date=2016-11-01&rft.pub=IEEE&rft.eissn=2164-0580&rft.spage=629&rft.epage=636&rft_id=info:doi/10.1109%2FHUMANOIDS.2016.7803340&rft.externalDocID=7803340