Using One-Class SVMs and MP for Audio Recognition of Action Scenes

Author

FengJuan, Guo ; ShuQian, Shan ; Xiaohui, Wang

Author_Institution

Sch. of Sci. & Technol., North China Electr. Power Univ., Baoding, China

Volume

1

fYear

2010

fDate

6-7 March 2010

Firstpage

401

Lastpage

404

Abstract

The paper presents a system for sounds classification of action scenes. We propose to use the matching pursuit (MP) algorithm to obtain effective time-frequency features. The MP-based method utilizes a dictionary of atoms for feature selection, resulting in a flexible, intuitive and physically interpretable set of features. Then apply one-class support vector machines (1-SVMs) together with MP features and MFCC to classify the sounds. The paper considers six type sounds: the sword sound, the club sound, the unarmed sound, the broken sound, the metal-falling sound and the shout sound. Experimental results prove that the method is effective.

Keywords

audio signal processing; feature extraction; signal classification; support vector machines; time-frequency analysis; action scenes; audio recognition; broken sound; club sound; feature selection; matching pursuit algorithm; metal-falling sound; one-class SVM; shout sound; sounds classification; support vector machines; sword sound; time-frequency features; unarmed sound; Art; Computer science; Dictionaries; Educational technology; Layout; Matching pursuit algorithms; Motion pictures; Speech recognition; Support vector machine classification; Support vector machines; Gabor atom; audio feature; matching pursuit; one-class SVM;

fLanguage

English

Publisher

ieee

Conference_Titel

Education Technology and Computer Science (ETCS), 2010 Second International Workshop on

Conference_Location

Wuhan

Print_ISBN

978-1-4244-6388-6

Electronic_ISBN

978-1-4244-6389-3

Type

conf

DOI

10.1109/ETCS.2010.164

Filename

5459014