A study on voiceprint identification systems under the Internet environment

Author

Yang, Hai-Yan ; Jing, Xin-Xing

Author_Institution

Sch. of Inf. & Commun., Guilin Univ. of Electron. Technol., Guilin, China

fYear

2011

fDate

10-13 July 2011

Firstpage

171

Lastpage

176

Abstract

A system of voiceprint identification under the Internet environment is discussed in this paper. Considering the limited storages of clients and the features of network transmission, Client/Server model is adopted in the system. In the client-side, feature parameters are extracted. This paper analyzes the performances with different parameters, and a new parameter (WPTMFC) and a mixed parameter under fisher criterion are given. After comparing recognition model of GMM and LSSVM, we choose LSSVM for the server and the process of recognition is performed. The final results show that the system has good recognition rate and is feasible for the practical applications.

Keywords

Internet; client-server systems; feature extraction; least squares approximations; speech recognition; support vector machines; GMM recognition model; Gaussian mixture model; Internet environment; LSSVM recognition model; client-server model; feature parameter extraction; fisher criterion; least squares support vector machine; network transmission; voiceprint identification systems; Feature extraction; Servers; Spectrogram; Speech; Speech recognition; Training; Client/Server model; GMM; LSSVM; WPTMFC;

fLanguage

English

Publisher

ieee

Conference_Titel

Wavelet Analysis and Pattern Recognition (ICWAPR), 2011 International Conference on

Conference_Location

Guilin

ISSN

2158-5695

Print_ISBN

978-1-4577-0283-9

Type

conf

DOI

10.1109/ICWAPR.2011.6014487

Filename

6014487