Feature analysis and extraction for audio automatic classification

Author

Liang, Bai ; Hu Yaali ; Songyang, Lao ; Jianyun, Chen ; Lingda, Wu

Author_Institution

Multimedia R&D Center, Nat. Univ. of Defense & Technol., ChangSha, China

Volume

1

fYear

2005

fDate

10-12 Oct. 2005

Firstpage

767

Abstract

Feature analysis and extraction are the foundation of audio automatic classification. This paper divides audio streams into five classes: silence, noise, pure speech, speech over background sound and music. We present our work on audio feature analysis and extraction on the frame level and clip level. Four new features are proposed, including silence ratio, pitch frequency standard deviation, harmonicity ratio and smooth pitch ratio. We have presented an SVM based approach to classification. The effectiveness of the features is evaluated in experiments. Experiment results show that the features we selected and proposed are rational and effective.

Keywords

audio signal processing; feature extraction; pattern classification; speech processing; support vector machines; SVM classification; audio automatic classification; audio clip level; audio feature analysis; audio feature extraction; audio frame level; content based audio classification; harmonicity ratio; noise audio stream; pitch frequency standard deviation; pure speech audio stream; silence audio stream; silence ratio; smooth pitch ratio; support vector machine; Acoustic noise; Background noise; Data mining; Feature extraction; Information analysis; Music; Speech enhancement; Streaming media; Support vector machine classification; Support vector machines; Feature analysis and extraction; content-based audio classification; support vector machines;

fLanguage

English

Publisher

ieee

Conference_Titel

Systems, Man and Cybernetics, 2005 IEEE International Conference on

Print_ISBN

0-7803-9298-1

Type

conf

DOI

10.1109/ICSMC.2005.1571239

Filename

1571239