Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning
Autor(a) principal: | |
---|---|
Data de Publicação: | 2015 |
Outros Autores: | , , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/1822/39108 |
Resumo: | There is currently an increasing demand for robots able to acquire the sequential organization of tasks from social learning interactions with ordinary people. Interactive learning-by-demonstration and communication is a promising research topic in current robotics research. However, the efficient acquisition of generalized task representations that allow the robot to adapt to different users and contexts is a major challenge. In this paper, we present a dynamic neural field (DNF) model that is inspired by the hypothesis that the nervous system uses the off-line re-activation of initial memory traces to incrementally incorporate new information into structured knowledge. To achieve this, the model combines fast activation-based learning to robustly represent sequential information from single task demonstrations with slower, weight-based learning during internal simulations to establish longer-term associations between neural populations representing individual subtasks. The efficiency of the learning process is tested in an assembly paradigm in which the humanoid robot ARoS learns to construct a toy vehicle from its parts. User demonstrations with different serial orders together with the correction of initial prediction errors allow the robot to acquire generalized task knowledge about possible serial orders and the longer term dependencies between subgoals in very few social learning interactions. This success is shown in a joint action scenario in which ARoS uses the newly acquired assembly plan to construct the toy together with a human partner. |
id |
RCAP_891566ddb895f5f6315148e5ea1299c5 |
---|---|
oai_identifier_str |
oai:repositorium.sdum.uminho.pt:1822/39108 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learningOff-line learningAdaptive robotSequential taskDynamic neural field modelPersistent neural activitySocial learningDynamic neural fieldPersistent activityCiências Naturais::MatemáticasScience & TechnologyThere is currently an increasing demand for robots able to acquire the sequential organization of tasks from social learning interactions with ordinary people. Interactive learning-by-demonstration and communication is a promising research topic in current robotics research. However, the efficient acquisition of generalized task representations that allow the robot to adapt to different users and contexts is a major challenge. In this paper, we present a dynamic neural field (DNF) model that is inspired by the hypothesis that the nervous system uses the off-line re-activation of initial memory traces to incrementally incorporate new information into structured knowledge. To achieve this, the model combines fast activation-based learning to robustly represent sequential information from single task demonstrations with slower, weight-based learning during internal simulations to establish longer-term associations between neural populations representing individual subtasks. The efficiency of the learning process is tested in an assembly paradigm in which the humanoid robot ARoS learns to construct a toy vehicle from its parts. User demonstrations with different serial orders together with the correction of initial prediction errors allow the robot to acquire generalized task knowledge about possible serial orders and the longer term dependencies between subgoals in very few social learning interactions. This success is shown in a joint action scenario in which ARoS uses the newly acquired assembly plan to construct the toy together with a human partner.The work was funded by FCT - Fundacao para a Ciencia e Tecnologia, through the PhD Grants SFRH/BD/48529/2008 and SFRH/BD/41179/2007 and Project NETT: Neural Engineering Transformative Technologies, EU-FP7 ITN (nr. 289146) and the FCT-Research Center CMAT (PEst-OE/MAT/UI0013/2014).ElsevierUniversidade do MinhoSousa, Emanuel Augusto FreitasErlhagen, WolframFerreira, Flora José RochaBicho, Estela2015-022015-02-01T00:00:00Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleapplication/pdfhttp://hdl.handle.net/1822/39108eng0893-608010.1016/j.neunet.2015.09.00226548945http://www.sciencedirect.com/science/article/pii/S089360801500177Xinfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2023-07-21T12:54:30Zoai:repositorium.sdum.uminho.pt:1822/39108Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T19:54:04.652845Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning |
title |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning |
spellingShingle |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning Sousa, Emanuel Augusto Freitas Off-line learning Adaptive robot Sequential task Dynamic neural field model Persistent neural activity Social learning Dynamic neural field Persistent activity Ciências Naturais::Matemáticas Science & Technology |
title_short |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning |
title_full |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning |
title_fullStr |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning |
title_full_unstemmed |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning |
title_sort |
Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning |
author |
Sousa, Emanuel Augusto Freitas |
author_facet |
Sousa, Emanuel Augusto Freitas Erlhagen, Wolfram Ferreira, Flora José Rocha Bicho, Estela |
author_role |
author |
author2 |
Erlhagen, Wolfram Ferreira, Flora José Rocha Bicho, Estela |
author2_role |
author author author |
dc.contributor.none.fl_str_mv |
Universidade do Minho |
dc.contributor.author.fl_str_mv |
Sousa, Emanuel Augusto Freitas Erlhagen, Wolfram Ferreira, Flora José Rocha Bicho, Estela |
dc.subject.por.fl_str_mv |
Off-line learning Adaptive robot Sequential task Dynamic neural field model Persistent neural activity Social learning Dynamic neural field Persistent activity Ciências Naturais::Matemáticas Science & Technology |
topic |
Off-line learning Adaptive robot Sequential task Dynamic neural field model Persistent neural activity Social learning Dynamic neural field Persistent activity Ciências Naturais::Matemáticas Science & Technology |
description |
There is currently an increasing demand for robots able to acquire the sequential organization of tasks from social learning interactions with ordinary people. Interactive learning-by-demonstration and communication is a promising research topic in current robotics research. However, the efficient acquisition of generalized task representations that allow the robot to adapt to different users and contexts is a major challenge. In this paper, we present a dynamic neural field (DNF) model that is inspired by the hypothesis that the nervous system uses the off-line re-activation of initial memory traces to incrementally incorporate new information into structured knowledge. To achieve this, the model combines fast activation-based learning to robustly represent sequential information from single task demonstrations with slower, weight-based learning during internal simulations to establish longer-term associations between neural populations representing individual subtasks. The efficiency of the learning process is tested in an assembly paradigm in which the humanoid robot ARoS learns to construct a toy vehicle from its parts. User demonstrations with different serial orders together with the correction of initial prediction errors allow the robot to acquire generalized task knowledge about possible serial orders and the longer term dependencies between subgoals in very few social learning interactions. This success is shown in a joint action scenario in which ARoS uses the newly acquired assembly plan to construct the toy together with a human partner. |
publishDate |
2015 |
dc.date.none.fl_str_mv |
2015-02 2015-02-01T00:00:00Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/1822/39108 |
url |
http://hdl.handle.net/1822/39108 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
0893-6080 10.1016/j.neunet.2015.09.002 26548945 http://www.sciencedirect.com/science/article/pii/S089360801500177X |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.publisher.none.fl_str_mv |
Elsevier |
publisher.none.fl_str_mv |
Elsevier |
dc.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799133138920669184 |