Traditional low-rank feature lose the temporal information among action sequence. To obtain the temporal information, we split an action video into multiple action subsequences and concatenate all the low-rank features of subsequences according to their time order. Then we recognize actions by learning a novel dictionary model from concatenated low-rank features. However, traditional dictionary learning models usually neglect the similarity among the coding coefficients and have bad performance in dealing with non-linearly separable data. To overcome these shortcomings, we present a novel similarity constrained discriminative kernel dictionary learning for action recognition. The effectiveness of the proposed method is verified on three benchmarks, and the experimental results show the promising results of our method for action recognition.
Shijian HUANG
Chongqing University,Yangtze Normal University
Junyong YE
Chongqing University
Tongqing WANG
Chongqing University
Li JIANG
Chongqing University of Posts and Telecommunications
Changyuan XING
Yangtze Normal University
Yang LI
Chongqing University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Shijian HUANG, Junyong YE, Tongqing WANG, Li JIANG, Changyuan XING, Yang LI, "Learning a Similarity Constrained Discriminative Kernel Dictionary from Concatenated Low-Rank Features for Action Recognition" in IEICE TRANSACTIONS on Information,
vol. E99-D, no. 2, pp. 541-544, February 2016, doi: 10.1587/transinf.2015EDL8148.
Abstract: Traditional low-rank feature lose the temporal information among action sequence. To obtain the temporal information, we split an action video into multiple action subsequences and concatenate all the low-rank features of subsequences according to their time order. Then we recognize actions by learning a novel dictionary model from concatenated low-rank features. However, traditional dictionary learning models usually neglect the similarity among the coding coefficients and have bad performance in dealing with non-linearly separable data. To overcome these shortcomings, we present a novel similarity constrained discriminative kernel dictionary learning for action recognition. The effectiveness of the proposed method is verified on three benchmarks, and the experimental results show the promising results of our method for action recognition.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2015EDL8148/_p
Copy
@ARTICLE{e99-d_2_541,
author={Shijian HUANG, Junyong YE, Tongqing WANG, Li JIANG, Changyuan XING, Yang LI, },
journal={IEICE TRANSACTIONS on Information},
title={Learning a Similarity Constrained Discriminative Kernel Dictionary from Concatenated Low-Rank Features for Action Recognition},
year={2016},
volume={E99-D},
number={2},
pages={541-544},
abstract={Traditional low-rank feature lose the temporal information among action sequence. To obtain the temporal information, we split an action video into multiple action subsequences and concatenate all the low-rank features of subsequences according to their time order. Then we recognize actions by learning a novel dictionary model from concatenated low-rank features. However, traditional dictionary learning models usually neglect the similarity among the coding coefficients and have bad performance in dealing with non-linearly separable data. To overcome these shortcomings, we present a novel similarity constrained discriminative kernel dictionary learning for action recognition. The effectiveness of the proposed method is verified on three benchmarks, and the experimental results show the promising results of our method for action recognition.},
keywords={},
doi={10.1587/transinf.2015EDL8148},
ISSN={1745-1361},
month={February},}
Copy
TY - JOUR
TI - Learning a Similarity Constrained Discriminative Kernel Dictionary from Concatenated Low-Rank Features for Action Recognition
T2 - IEICE TRANSACTIONS on Information
SP - 541
EP - 544
AU - Shijian HUANG
AU - Junyong YE
AU - Tongqing WANG
AU - Li JIANG
AU - Changyuan XING
AU - Yang LI
PY - 2016
DO - 10.1587/transinf.2015EDL8148
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E99-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2016
AB - Traditional low-rank feature lose the temporal information among action sequence. To obtain the temporal information, we split an action video into multiple action subsequences and concatenate all the low-rank features of subsequences according to their time order. Then we recognize actions by learning a novel dictionary model from concatenated low-rank features. However, traditional dictionary learning models usually neglect the similarity among the coding coefficients and have bad performance in dealing with non-linearly separable data. To overcome these shortcomings, we present a novel similarity constrained discriminative kernel dictionary learning for action recognition. The effectiveness of the proposed method is verified on three benchmarks, and the experimental results show the promising results of our method for action recognition.
ER -