Context Information Fusion Method for Temporal Action Proposals
In the field of human activity localization and recognition in videos, the existing temporal action proposal methods have not solved the long-term dependence problem better, which results in lower recall rates of proposals. In view of this problem, a method based on context information fusion for te...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press
2021-03-01
|
| Series: | Jisuanji kexue yu tansuo |
| Subjects: | |
| Online Access: | http://fcst.ceaj.org/CN/abstract/abstract2596.shtml |
| _version_ | 1824010655205687296 |
|---|---|
| author | WANG Xinwen, XIE Linbo, PENG Li |
| author_facet | WANG Xinwen, XIE Linbo, PENG Li |
| author_sort | WANG Xinwen, XIE Linbo, PENG Li |
| collection | DOAJ |
| description | In the field of human activity localization and recognition in videos, the existing temporal action proposal methods have not solved the long-term dependence problem better, which results in lower recall rates of proposals. In view of this problem, a method based on context information fusion for temporal action proposals is proposed in this paper. Firstly, the spatiotemporal features of video units are extracted by the 3D convolutional network. Then, the bidirectional recurrent network is used to construct the context relationship for predicting the temporal action proposals. Considering the problems of more parameters and the vanishing gradient in the gated recurrent unit (GRU), a simplified-GRU (S-GRU) is proposed, in which the input features control the gating structure to enhance the parallel computing capability and the weighted average is introduced to enhance the ability of the gated recurrent unit to adaptively fuse the history and current time information. Finally, experimental results on the Thumos14 dataset demonstrate that the method based on the bidirectional S-GRU for temporal action proposals improves the recall rate of proposals. |
| first_indexed | 2024-12-18T21:26:21Z |
| format | Article |
| id | doaj.art-7881ae4d7ed241e7a7aaf32c7c23c617 |
| institution | Directory Open Access Journal |
| issn | 1673-9418 |
| language | zho |
| last_indexed | 2024-12-18T21:26:21Z |
| publishDate | 2021-03-01 |
| publisher | Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press |
| record_format | Article |
| series | Jisuanji kexue yu tansuo |
| spelling | doaj.art-7881ae4d7ed241e7a7aaf32c7c23c6172022-12-21T20:51:30ZzhoJournal of Computer Engineering and Applications Beijing Co., Ltd., Science PressJisuanji kexue yu tansuo1673-94182021-03-0115348649410.3778/j.issn.1673-9418.1912044Context Information Fusion Method for Temporal Action ProposalsWANG Xinwen, XIE Linbo, PENG Li0Engineering Research Center of Internet of Things Technology Applications (School of Internet of Things Engineering, Jiangnan University), Ministry of Education, Wuxi, Jiangsu 214122, ChinaIn the field of human activity localization and recognition in videos, the existing temporal action proposal methods have not solved the long-term dependence problem better, which results in lower recall rates of proposals. In view of this problem, a method based on context information fusion for temporal action proposals is proposed in this paper. Firstly, the spatiotemporal features of video units are extracted by the 3D convolutional network. Then, the bidirectional recurrent network is used to construct the context relationship for predicting the temporal action proposals. Considering the problems of more parameters and the vanishing gradient in the gated recurrent unit (GRU), a simplified-GRU (S-GRU) is proposed, in which the input features control the gating structure to enhance the parallel computing capability and the weighted average is introduced to enhance the ability of the gated recurrent unit to adaptively fuse the history and current time information. Finally, experimental results on the Thumos14 dataset demonstrate that the method based on the bidirectional S-GRU for temporal action proposals improves the recall rate of proposals.http://fcst.ceaj.org/CN/abstract/abstract2596.shtmlgated recurrent network (gru)vanishing gradientcontext informationtemporal action proposalstemporal action detection |
| spellingShingle | WANG Xinwen, XIE Linbo, PENG Li Context Information Fusion Method for Temporal Action Proposals Jisuanji kexue yu tansuo gated recurrent network (gru) vanishing gradient context information temporal action proposals temporal action detection |
| title | Context Information Fusion Method for Temporal Action Proposals |
| title_full | Context Information Fusion Method for Temporal Action Proposals |
| title_fullStr | Context Information Fusion Method for Temporal Action Proposals |
| title_full_unstemmed | Context Information Fusion Method for Temporal Action Proposals |
| title_short | Context Information Fusion Method for Temporal Action Proposals |
| title_sort | context information fusion method for temporal action proposals |
| topic | gated recurrent network (gru) vanishing gradient context information temporal action proposals temporal action detection |
| url | http://fcst.ceaj.org/CN/abstract/abstract2596.shtml |
| work_keys_str_mv | AT wangxinwenxielinbopengli contextinformationfusionmethodfortemporalactionproposals |