Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields
Autor(a) principal: | |
---|---|
Data de Publicação: | 2013 |
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/10174/10646 https://doi.org/10.1007/978-3-642-38628-2_9 |
Resumo: | This paper presents a new algorithm for trajectory classifi- cation of human activities. The presented framework uses a mixture of parametric space-variant vector fields to describe pedestrian’s trajecto- ries. An advantage of the proposed method is that the vector fields are not constant and depend on the pedestrian’s localization. This means that the switching motion among vector fields may occur at any image location and should be accurately estimated. In this paper, the model is equipped with a novel methodology to estimate the switching probabilities among motion regimes. More specifically, we propose an iterative optimization of switching probabilities based on the natural gradient vector, with respect to the Fisher information metric. This approach follows an information geometric framework and contrasts with more traditional approaches of constrained optimization in which euclidean gradient based methods are used combined with probability simplex constraints. We testify the per- formance superiority of the proposed approach in the classification of pedestrian’s trajectories in synthetic and real data sets concerning farfield surveillance scenarios. |
id |
RCAP_afb28a6a0622bc5f7dd46a8bc90224a3 |
---|---|
oai_identifier_str |
oai:dspace.uevora.pt:10174/10646 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector FieldsThis paper presents a new algorithm for trajectory classifi- cation of human activities. The presented framework uses a mixture of parametric space-variant vector fields to describe pedestrian’s trajecto- ries. An advantage of the proposed method is that the vector fields are not constant and depend on the pedestrian’s localization. This means that the switching motion among vector fields may occur at any image location and should be accurately estimated. In this paper, the model is equipped with a novel methodology to estimate the switching probabilities among motion regimes. More specifically, we propose an iterative optimization of switching probabilities based on the natural gradient vector, with respect to the Fisher information metric. This approach follows an information geometric framework and contrasts with more traditional approaches of constrained optimization in which euclidean gradient based methods are used combined with probability simplex constraints. We testify the per- formance superiority of the proposed approach in the classification of pedestrian’s trajectories in synthetic and real data sets concerning farfield surveillance scenarios.Springer Berlin Heidelberg2014-02-07T14:32:20Z2014-02-072013-01-01T00:00:00Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articlehttp://hdl.handle.net/10174/10646http://hdl.handle.net/10174/10646https://doi.org/10.1007/978-3-642-38628-2_9engNascimento, Jacinto C.; Barão, Miguel; Marques, Jorge S.; Lemos, João M.Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields, In Pattern Recognition and Image Analysis, 79-88, ISBN: 978-3-642-38627-5. Berlin, Heidelberg: Springer Berlin Heidelberg, 2013.ndndndndNascimento, Jacinto C.Barão, MiguelMarques, Jorge S.Lemos, João M.info: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:RCAAP2024-01-03T18:53:39Zoai:dspace.uevora.pt:10174/10646Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-20T01:04:30.249825Repositório 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 |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields |
title |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields |
spellingShingle |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields Nascimento, Jacinto C. |
title_short |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields |
title_full |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields |
title_fullStr |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields |
title_full_unstemmed |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields |
title_sort |
Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields |
author |
Nascimento, Jacinto C. |
author_facet |
Nascimento, Jacinto C. Barão, Miguel Marques, Jorge S. Lemos, João M. |
author_role |
author |
author2 |
Barão, Miguel Marques, Jorge S. Lemos, João M. |
author2_role |
author author author |
dc.contributor.author.fl_str_mv |
Nascimento, Jacinto C. Barão, Miguel Marques, Jorge S. Lemos, João M. |
description |
This paper presents a new algorithm for trajectory classifi- cation of human activities. The presented framework uses a mixture of parametric space-variant vector fields to describe pedestrian’s trajecto- ries. An advantage of the proposed method is that the vector fields are not constant and depend on the pedestrian’s localization. This means that the switching motion among vector fields may occur at any image location and should be accurately estimated. In this paper, the model is equipped with a novel methodology to estimate the switching probabilities among motion regimes. More specifically, we propose an iterative optimization of switching probabilities based on the natural gradient vector, with respect to the Fisher information metric. This approach follows an information geometric framework and contrasts with more traditional approaches of constrained optimization in which euclidean gradient based methods are used combined with probability simplex constraints. We testify the per- formance superiority of the proposed approach in the classification of pedestrian’s trajectories in synthetic and real data sets concerning farfield surveillance scenarios. |
publishDate |
2013 |
dc.date.none.fl_str_mv |
2013-01-01T00:00:00Z 2014-02-07T14:32:20Z 2014-02-07 |
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/10174/10646 http://hdl.handle.net/10174/10646 https://doi.org/10.1007/978-3-642-38628-2_9 |
url |
http://hdl.handle.net/10174/10646 https://doi.org/10.1007/978-3-642-38628-2_9 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
Nascimento, Jacinto C.; Barão, Miguel; Marques, Jorge S.; Lemos, João M.Efficient Optimization Algorithm for Space-Variant Mixture of Vector Fields, In Pattern Recognition and Image Analysis, 79-88, ISBN: 978-3-642-38627-5. Berlin, Heidelberg: Springer Berlin Heidelberg, 2013. nd nd nd nd |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.publisher.none.fl_str_mv |
Springer Berlin Heidelberg |
publisher.none.fl_str_mv |
Springer Berlin Heidelberg |
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_ |
1799136528759259136 |