Representation for Action Recognition Using Trajectory-Based Low-Level Local Feature and Mid-Level Motion Feature
Joint Authors
Li, Xiaoqiang
Wang, Dan
Zhang, Yin
Source
Applied Computational Intelligence and Soft Computing
Issue
Vol. 2017, Issue 2017 (31 Dec. 2017), pp.1-7, 7 p.
Publisher
Hindawi Publishing Corporation
Publication Date
2017-10-19
Country of Publication
Egypt
No. of Pages
7
Main Subjects
Information Technology and Computer Science
Abstract EN
The dense trajectories and low-level local features are widely used in action recognition recently.
However, most of these methods ignore the motion part of action which is the key factor to distinguish the different human action.
This paper proposes a new two-layer model of representation for action recognition by describing the video with low-level features and mid-level motion part model.
Firstly, we encode the compensated flow (w-flow) trajectory-based local features with Fisher Vector (FV) to retain the low-level characteristic of motion.
Then, the motion parts are extracted by clustering the similar trajectories with spatiotemporal distance between trajectories.
Finally the representation for action video is the concatenation of low-level descriptors encoding vector and motion part encoding vector.
It is used as input to the LibSVM for action recognition.
The experiment results demonstrate the improvements on J-HMDB and YouTube datasets, which obtain 67.4% and 87.6%, respectively.
American Psychological Association (APA)
Li, Xiaoqiang& Wang, Dan& Zhang, Yin. 2017. Representation for Action Recognition Using Trajectory-Based Low-Level Local Feature and Mid-Level Motion Feature. Applied Computational Intelligence and Soft Computing،Vol. 2017, no. 2017, pp.1-7.
https://search.emarefa.net/detail/BIM-1121427
Modern Language Association (MLA)
Li, Xiaoqiang…[et al.]. Representation for Action Recognition Using Trajectory-Based Low-Level Local Feature and Mid-Level Motion Feature. Applied Computational Intelligence and Soft Computing No. 2017 (2017), pp.1-7.
https://search.emarefa.net/detail/BIM-1121427
American Medical Association (AMA)
Li, Xiaoqiang& Wang, Dan& Zhang, Yin. Representation for Action Recognition Using Trajectory-Based Low-Level Local Feature and Mid-Level Motion Feature. Applied Computational Intelligence and Soft Computing. 2017. Vol. 2017, no. 2017, pp.1-7.
https://search.emarefa.net/detail/BIM-1121427
Data Type
Journal Articles
Language
English
Notes
Includes bibliographical references
Record ID
BIM-1121427