Structured action classification with hypergraph regularization

Chaoqun Hong; Jun Yu; Xuhui Chen

doi:10.1109/smc.2014.6974362

Structured action classification with hypergraph regularization

Chaoqun Hong
, Jun Yu
, Xuhui Chen

Xiamen University of Technology
Hangzhou Dianzi University

Research output: Contribution to journal › Conference article › peer-review

1 Scopus citations

Abstract

Traditional multi-class classifying methods treat outputs separately. It leads to a multiclass problem with a very large number of classes and downgrades the performance of classifiers. Actually, the outputs of different testing samples are usually interdependent. Therefore, we propose a novel method of structured classification based on SVM and hypergraph regularization (Hyper-SSVM). First, it exploits the structure and dependencies within classifying outputs. Second, we impose local constraints to samples by using Hypergraph regularization. We apply the proposed Hyper-SSVM to action classification. The experimental results demonstrate the effectiveness of the proposed method.

Original language	English
Article number	6974362
Pages (from-to)	2853-2858
Number of pages	6
Journal	Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics
Volume	2014-January
Issue number	January
DOIs	https://doi.org/10.1109/smc.2014.6974362
State	Published - 2014
Externally published	Yes
Event	2014 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2014 - San Diego, United States Duration: 5 Oct 2014 → 8 Oct 2014

Keywords

Hypergraph regularization
Motion classification
Sturctured SVM

Access to Document

10.1109/smc.2014.6974362

Cite this

@article{4cf282202815473f81e506575bd651e6,

title = "Structured action classification with hypergraph regularization",

abstract = "Traditional multi-class classifying methods treat outputs separately. It leads to a multiclass problem with a very large number of classes and downgrades the performance of classifiers. Actually, the outputs of different testing samples are usually interdependent. Therefore, we propose a novel method of structured classification based on SVM and hypergraph regularization (Hyper-SSVM). First, it exploits the structure and dependencies within classifying outputs. Second, we impose local constraints to samples by using Hypergraph regularization. We apply the proposed Hyper-SSVM to action classification. The experimental results demonstrate the effectiveness of the proposed method.",

keywords = "Hypergraph regularization, Motion classification, Sturctured SVM",

author = "Chaoqun Hong and Jun Yu and Xuhui Chen",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2014 ; Conference date: 05-10-2014 Through 08-10-2014",

year = "2014",

doi = "10.1109/smc.2014.6974362",

language = "英语",

volume = "2014-January",

pages = "2853--2858",

journal = "Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics",

issn = "1062-922X",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "January",

}

TY - JOUR

T1 - Structured action classification with hypergraph regularization

AU - Hong, Chaoqun

AU - Yu, Jun

AU - Chen, Xuhui

PY - 2014

Y1 - 2014

N2 - Traditional multi-class classifying methods treat outputs separately. It leads to a multiclass problem with a very large number of classes and downgrades the performance of classifiers. Actually, the outputs of different testing samples are usually interdependent. Therefore, we propose a novel method of structured classification based on SVM and hypergraph regularization (Hyper-SSVM). First, it exploits the structure and dependencies within classifying outputs. Second, we impose local constraints to samples by using Hypergraph regularization. We apply the proposed Hyper-SSVM to action classification. The experimental results demonstrate the effectiveness of the proposed method.

AB - Traditional multi-class classifying methods treat outputs separately. It leads to a multiclass problem with a very large number of classes and downgrades the performance of classifiers. Actually, the outputs of different testing samples are usually interdependent. Therefore, we propose a novel method of structured classification based on SVM and hypergraph regularization (Hyper-SSVM). First, it exploits the structure and dependencies within classifying outputs. Second, we impose local constraints to samples by using Hypergraph regularization. We apply the proposed Hyper-SSVM to action classification. The experimental results demonstrate the effectiveness of the proposed method.

KW - Hypergraph regularization

KW - Motion classification

KW - Sturctured SVM

UR - https://www.scopus.com/pages/publications/84938153554

U2 - 10.1109/smc.2014.6974362

DO - 10.1109/smc.2014.6974362

M3 - 会议文章

AN - SCOPUS:84938153554

SN - 1062-922X

VL - 2014-January

SP - 2853

EP - 2858

JO - Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics

JF - Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics

IS - January

M1 - 6974362

T2 - 2014 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2014

Y2 - 5 October 2014 through 8 October 2014

ER -

Structured action classification with hypergraph regularization

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this