Low-rank representation based action recognition

Xiangrong Zhang; Yang Yang; Hanghua Jia; Huiyu Zhou; Licheng Jiao

Low-rank representation based action recognition

Xiangrong Zhang, Yang Yang, Hanghua Jia, Huiyu Zhou, Licheng Jiao

Research output: Contribution to conference › Paper › peer-review

5 Citations (Scopus)

276 Downloads (Pure)

Abstract

Human action recognition is an important problem in computer vision, which has been applied to many applications. However, how to learn an accurate and discriminative representation of videos based on the features extracted from videos still remains to be a challenging problem. In this paper, we propose a novel method named low-rank representation based action recognition to recognize human actions. Given a dictionary, low-rank representation aims at finding the lowestrank representation of all data, which can capture the global data structures. According to its characteristics, low-rank representation is robust against noises. Experimental results demonstrate the effectiveness of the proposed approach on several publicly available datasets.

Original language	English
Number of pages	7
Publication status	Published - 2014
Event	2014 International Joint Conference on Neural Networks - Beijing International Convention Center, 8 Beichen East Road, Chaoyang District, Beijing, China, Beijing, China Duration: 06 Jul 2014 → 11 Jul 2014

Conference

Conference	2014 International Joint Conference on Neural Networks
Abbreviated title	IJCNN 2014
Country/Territory	China
City	Beijing
Period	06/07/2014 → 11/07/2014
Other	International Joint Conference on Neural Networks is the largest technical event in the field of neural networks, jointly organized by IEEE Computational Intelligence Society and Intenational Neural Network Society. In 2014, International Joint Conference on Neural Networks will be part of the 2104 IEEE World Congress on Computational Intelligence.

Access to Document

Low-rank Representation Based Action RecognitionSubmitted manuscript, 551 KB

1 Finished

R1118ECI: Centre for Secure Information Technologies (CSIT)
McCanny, J. V., Cowan, C., Crookes, D., Fusco, V., Linton, D., Liu, W., Miller, P., O'Neill, M., Scanlon, W. & Sezer, S.
01/08/2009 → 30/06/2014
Project: Research

Cite this

@conference{fe476e09b08d41b0b81575b9169131f1,

title = "Low-rank representation based action recognition",

abstract = "Human action recognition is an important problem in computer vision, which has been applied to many applications. However, how to learn an accurate and discriminative representation of videos based on the features extracted from videos still remains to be a challenging problem. In this paper, we propose a novel method named low-rank representation based action recognition to recognize human actions. Given a dictionary, low-rank representation aims at finding the lowestrank representation of all data, which can capture the global data structures. According to its characteristics, low-rank representation is robust against noises. Experimental results demonstrate the effectiveness of the proposed approach on several publicly available datasets.",

author = "Xiangrong Zhang and Yang Yang and Hanghua Jia and Huiyu Zhou and Licheng Jiao",

year = "2014",

language = "English",

note = "2014 International Joint Conference on Neural Networks, IJCNN 2014 ; Conference date: 06-07-2014 Through 11-07-2014",

}

TY - CONF

T1 - Low-rank representation based action recognition

AU - Zhang, Xiangrong

AU - Yang, Yang

AU - Jia, Hanghua

AU - Zhou, Huiyu

AU - Jiao, Licheng

PY - 2014

Y1 - 2014

N2 - Human action recognition is an important problem in computer vision, which has been applied to many applications. However, how to learn an accurate and discriminative representation of videos based on the features extracted from videos still remains to be a challenging problem. In this paper, we propose a novel method named low-rank representation based action recognition to recognize human actions. Given a dictionary, low-rank representation aims at finding the lowestrank representation of all data, which can capture the global data structures. According to its characteristics, low-rank representation is robust against noises. Experimental results demonstrate the effectiveness of the proposed approach on several publicly available datasets.

AB - Human action recognition is an important problem in computer vision, which has been applied to many applications. However, how to learn an accurate and discriminative representation of videos based on the features extracted from videos still remains to be a challenging problem. In this paper, we propose a novel method named low-rank representation based action recognition to recognize human actions. Given a dictionary, low-rank representation aims at finding the lowestrank representation of all data, which can capture the global data structures. According to its characteristics, low-rank representation is robust against noises. Experimental results demonstrate the effectiveness of the proposed approach on several publicly available datasets.

M3 - Paper

T2 - 2014 International Joint Conference on Neural Networks

Y2 - 6 July 2014 through 11 July 2014

ER -

Low-rank representation based action recognition

Abstract

Conference

Access to Document

Fingerprint

Projects

R1118ECI: Centre for Secure Information Technologies (CSIT)

Cite this