Efficient Model Learning from Joint-Action Demonstrations for Human-Robot Collaborative Tasks

We present a framework for automatically learning human user models from joint-action demonstrations that enables a robot to compute a robust policy for a collaborative task with a human. First, the demonstrated action sequences are clustered into different human types using an unsupervised learning...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Hri '15: ACM/IEEE International Conference on Human-Robot Interaction USB Stick s. 189 - 196
Hlavní autoři: Nikolaidis, Stefanos, Ramakrishnan, Ramya, Keren Gu, Shah, Julie
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: ACM 02.03.2015
Témata:
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract We present a framework for automatically learning human user models from joint-action demonstrations that enables a robot to compute a robust policy for a collaborative task with a human. First, the demonstrated action sequences are clustered into different human types using an unsupervised learning algorithm. A reward function is then learned for each type through the employment of an inverse reinforcement learning algorithm. The learned model is then incorporated into a mixed-observability Markov decision process (MOMDP) formulation, wherein the human type is a partially observable variable. With this framework, we can infer online the human type of a new user that was not included in the training set, and can compute a policy for the robot that will be aligned to the preference of this user. In a human subject experiment (n=30), participants agreed more strongly that the robot anticipated their actions when working with a robot incorporating the proposed framework (p<;0.01), compared to manually annotating robot actions. In trials where participants faced difficulty annotating the robot actions to complete the task, the proposed framework significantly improved team efficiency (p <;0.01). The robot incorporating the framework was also found to be more responsive to human actions compared to policies computed using a hand-coded reward function by a domain expert (p<;0.01). These results indicate that learning human user models from joint-action demonstrations and encoding them in a MOMDP formalism can support effective teaming in human-robot collaborative tasks.
AbstractList We present a framework for automatically learning human user models from joint-action demonstrations that enables a robot to compute a robust policy for a collaborative task with a human. First, the demonstrated action sequences are clustered into different human types using an unsupervised learning algorithm. A reward function is then learned for each type through the employment of an inverse reinforcement learning algorithm. The learned model is then incorporated into a mixed-observability Markov decision process (MOMDP) formulation, wherein the human type is a partially observable variable. With this framework, we can infer online the human type of a new user that was not included in the training set, and can compute a policy for the robot that will be aligned to the preference of this user. In a human subject experiment (n=30), participants agreed more strongly that the robot anticipated their actions when working with a robot incorporating the proposed framework (p<;0.01), compared to manually annotating robot actions. In trials where participants faced difficulty annotating the robot actions to complete the task, the proposed framework significantly improved team efficiency (p <;0.01). The robot incorporating the framework was also found to be more responsive to human actions compared to policies computed using a hand-coded reward function by a domain expert (p<;0.01). These results indicate that learning human user models from joint-action demonstrations and encoding them in a MOMDP formalism can support effective teaming in human-robot collaborative tasks.
Author Ramakrishnan, Ramya
Keren Gu
Nikolaidis, Stefanos
Shah, Julie
Author_xml – sequence: 1
  givenname: Stefanos
  surname: Nikolaidis
  fullname: Nikolaidis, Stefanos
  email: snikol@alum.mit.edu
  organization: Comput. Sci. & Artificial Intell. Lab., Massachusetts Inst. of Technol., Cambridge, MA, USA
– sequence: 2
  givenname: Ramya
  surname: Ramakrishnan
  fullname: Ramakrishnan, Ramya
  email: ramyaram@mit.edu
  organization: Comput. Sci. & Artificial Intell. Lab., Massachusetts Inst. of Technol., Cambridge, MA, USA
– sequence: 3
  surname: Keren Gu
  fullname: Keren Gu
  email: kgu@mit.edu
  organization: Comput. Sci. & Artificial Intell. Lab., Massachusetts Inst. of Technol., Cambridge, MA, USA
– sequence: 4
  givenname: Julie
  surname: Shah
  fullname: Shah, Julie
  email: julie_a_shah@csail.mit.edu
  organization: Comput. Sci. & Artificial Intell. Lab., Massachusetts Inst. of Technol., Cambridge, MA, USA
BookMark eNotjE9LwzAcQCMoqLNnD17yBTrzv8lxzLkpFUHmUUaa_irRNpEkCn57N-bp8XjwLtFpiAEQuqZkTqmQt0wZJaSYHylPUGUavQ-EM60ZPUdVzh-EEKa0lMxcoLfVMHjnIRT8FHsYcQs2BR_e8ZDihB-jD6VeuOJjwHcwxZBLsgfLeIgJb74nG-qX2MWCl3EcbRcP-Qfw1ubPfIXOBjtmqP45Q6_3q-1yU7fP64floq0tV7TUWjJGnG4EgGDO6I6S3hpJOwdag3HS9ES7BjrCB-WscZQ6K43hjVKCG8tn6Ob49QCw-0p-sul3t98SxRr-BwGRU_M
ContentType Conference Proceeding
DBID 6IE
6IL
CBEJK
RIE
RIL
DOI 10.1145/2696454.2696455
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume
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
EISBN 9781450328821
1450328822
EndPage 196
ExternalDocumentID 8520627
Genre orig-research
GroupedDBID 6IE
6IF
6IL
6IN
AAJGR
AAWTH
ADFMO
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
IEGSK
OCL
RIE
RIL
ID FETCH-LOGICAL-a361t-85220c874ee42c98b10da951bce88e9c59d08c7eb03f6ca9c11ca5993766439a3
IEDL.DBID RIE
ISICitedReferencesCount 149
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000371986600026&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 03:02:40 EDT 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed false
IsScholarly false
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-a361t-85220c874ee42c98b10da951bce88e9c59d08c7eb03f6ca9c11ca5993766439a3
PageCount 8
ParticipantIDs ieee_primary_8520627
PublicationCentury 2000
PublicationDate 2015-03-02
PublicationDateYYYYMMDD 2015-03-02
PublicationDate_xml – month: 03
  year: 2015
  text: 2015-03-02
  day: 02
PublicationDecade 2010
PublicationTitle Hri '15: ACM/IEEE International Conference on Human-Robot Interaction USB Stick
PublicationTitleAbbrev HRI
PublicationYear 2015
Publisher ACM
Publisher_xml – name: ACM
SSID ssj0002685529
Score 2.0669918
Snippet We present a framework for automatically learning human user models from joint-action demonstrations that enables a robot to compute a robust policy for a...
SourceID ieee
SourceType Publisher
StartPage 189
SubjectTerms Clustering algorithms
Collaboration
Computational modeling
Data models
Markov processes
Robots
Task analysis
Title Efficient Model Learning from Joint-Action Demonstrations for Human-Robot Collaborative Tasks
URI https://ieeexplore.ieee.org/document/8520627
WOSCitedRecordID wos000371986600026&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/eLvHCXMwlV3PS8MwFA6bePCksom_ycGj2do0aZOjzA3xMIZM2EVGfrzKUFtZu_39JmnZELx4SgiEhLyQ9-Ul3_cQuhORzt09QRIVWUmYVsKdg4kmzhckXv5KU8hDsolsOhWLhZx10P2OCwMA4fMZDHw1vOXb0mx8qGwoOPWqul3UzbKs4Wrt4ik0FZxT2ar3xIwPaSq9XNWgKfmv9CnBe0yO_zfuCervaXh4tnMwp6gDRQ-9jYPog-uAfR6zT9wqpL5jzxTBz-WqqMlDYCvgR_jy8K8xcoUdPsUhaE9eSl3WeLTfA1vAc1V9VH30OhnPR0-kzZFAVJLGNXFTo5ERGQNg1Eih48gqh5q0ASFAGi5tJEwGOkry1Chp4tgo7kFJ6rGISs7QQVEWcI5wImPLdc4lSxUDJrwwlbUOA0iTg-biAvX80iy_GxmMZbsql383X6Ejhy14-K5Fr9FBvd7ADTo023pVrW-D7X4AaqCblQ
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PS8MwFA5zCnpS2cTf5uDRbG2adMlR5mTqHEMm7CIjSV9lqKus3f5-k7RsCF48JQRCQl7I-_KS73sIXYtAp_aeIIkKEkmYVsKeg5Em1hdETv5KU0h9sonOcCgmEzmqoZs1FwYA_OczaLmqf8tPMrN0obK24NSp6m6hbc4YDUu21jqiQmPBOZWVfk_IeJvG0glWtcqS_0qg4v3H_f7_Rj5AzQ0RD4_WLuYQ1WDeQG89L_tgO2CXyewTVxqp79hxRfBjNpsX5NbzFfAdfDkAWJo5xxahYh-2Jy-Zzgrc3eyCFeCxyj_yJnq97427fVJlSSAqisOC2KnRwIgOA2DUSKHDIFEWN2kDQoA0XCaBMB3QQZTGRkkThkZxB0tih0ZUdITq82wOxwhHMky4TrlksWLAhJOmShKLAqRJQXNxghpuaabfpRDGtFqV07-br9Buf_w8mA4ehk9naM8iDe4_b9FzVC8WS7hAO2ZVzPLFpbfjD7DUntw
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=Hri+%2715%3A+ACM%2FIEEE+International+Conference+on+Human-Robot+Interaction+USB+Stick&rft.atitle=Efficient+Model+Learning+from+Joint-Action+Demonstrations+for+Human-Robot+Collaborative+Tasks&rft.au=Nikolaidis%2C+Stefanos&rft.au=Ramakrishnan%2C+Ramya&rft.au=Keren+Gu&rft.au=Shah%2C+Julie&rft.date=2015-03-02&rft.pub=ACM&rft.spage=189&rft.epage=196&rft_id=info:doi/10.1145%2F2696454.2696455&rft.externalDocID=8520627