Cargando…

Jointly Learning Multiple Sequential Dynamics for Human Action Recognition

Discovering visual dynamics during human actions is a challenging task for human action recognition. To deal with this problem, we theoretically propose the multi-task conditional random fields model and explore its application on human action recognition. For visual representation, we propose the p...

Descripción completa

Detalles Bibliográficos
Autores principales: Liu, An-An, Su, Yu-Ting, Nie, Wei-Zhi, Yang, Zhao-Xuan
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Public Library of Science 2015
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4493153/
https://www.ncbi.nlm.nih.gov/pubmed/26147979
http://dx.doi.org/10.1371/journal.pone.0130884
_version_ 1782379876550967296
author Liu, An-An
Su, Yu-Ting
Nie, Wei-Zhi
Yang, Zhao-Xuan
author_facet Liu, An-An
Su, Yu-Ting
Nie, Wei-Zhi
Yang, Zhao-Xuan
author_sort Liu, An-An
collection PubMed
description Discovering visual dynamics during human actions is a challenging task for human action recognition. To deal with this problem, we theoretically propose the multi-task conditional random fields model and explore its application on human action recognition. For visual representation, we propose the part-induced spatiotemporal action unit sequence to represent each action sample with multiple partwise sequential feature subspaces. For model learning, we propose the multi-task conditional random fields (MTCRFs) model to discover the sequence-specific structure and the sequence-shared relationship. Specifically, the multi-chain graph structure and the corresponding probabilistic model are designed to represent the interaction among multiple part-induced action unit sequences. Moreover we propose the model learning and inference methods to discover temporal context within individual action unit sequence and the latent correlation among different body parts. Extensive experiments are implemented to demonstrate the superiority of the proposed method on two popular RGB human action datasets, KTH & TJU, and the depth dataset in MSR Daily Activity 3D.
format Online
Article
Text
id pubmed-4493153
institution National Center for Biotechnology Information
language English
publishDate 2015
publisher Public Library of Science
record_format MEDLINE/PubMed
spelling pubmed-44931532015-07-15 Jointly Learning Multiple Sequential Dynamics for Human Action Recognition Liu, An-An Su, Yu-Ting Nie, Wei-Zhi Yang, Zhao-Xuan PLoS One Research Article Discovering visual dynamics during human actions is a challenging task for human action recognition. To deal with this problem, we theoretically propose the multi-task conditional random fields model and explore its application on human action recognition. For visual representation, we propose the part-induced spatiotemporal action unit sequence to represent each action sample with multiple partwise sequential feature subspaces. For model learning, we propose the multi-task conditional random fields (MTCRFs) model to discover the sequence-specific structure and the sequence-shared relationship. Specifically, the multi-chain graph structure and the corresponding probabilistic model are designed to represent the interaction among multiple part-induced action unit sequences. Moreover we propose the model learning and inference methods to discover temporal context within individual action unit sequence and the latent correlation among different body parts. Extensive experiments are implemented to demonstrate the superiority of the proposed method on two popular RGB human action datasets, KTH & TJU, and the depth dataset in MSR Daily Activity 3D. Public Library of Science 2015-07-06 /pmc/articles/PMC4493153/ /pubmed/26147979 http://dx.doi.org/10.1371/journal.pone.0130884 Text en © 2015 Liu et al http://creativecommons.org/licenses/by/4.0/ This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
spellingShingle Research Article
Liu, An-An
Su, Yu-Ting
Nie, Wei-Zhi
Yang, Zhao-Xuan
Jointly Learning Multiple Sequential Dynamics for Human Action Recognition
title Jointly Learning Multiple Sequential Dynamics for Human Action Recognition
title_full Jointly Learning Multiple Sequential Dynamics for Human Action Recognition
title_fullStr Jointly Learning Multiple Sequential Dynamics for Human Action Recognition
title_full_unstemmed Jointly Learning Multiple Sequential Dynamics for Human Action Recognition
title_short Jointly Learning Multiple Sequential Dynamics for Human Action Recognition
title_sort jointly learning multiple sequential dynamics for human action recognition
topic Research Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4493153/
https://www.ncbi.nlm.nih.gov/pubmed/26147979
http://dx.doi.org/10.1371/journal.pone.0130884
work_keys_str_mv AT liuanan jointlylearningmultiplesequentialdynamicsforhumanactionrecognition
AT suyuting jointlylearningmultiplesequentialdynamicsforhumanactionrecognition
AT nieweizhi jointlylearningmultiplesequentialdynamicsforhumanactionrecognition
AT yangzhaoxuan jointlylearningmultiplesequentialdynamicsforhumanactionrecognition