SVM-based speaker verification for coded and uncoded speech

Author

Janicki, Artur

Author_Institution

Inst. of Telecommun., Warsaw Univ. of Technol., Warsaw, Poland

fYear

2012

fDate

27-31 Aug. 2012

Firstpage

26

Lastpage

30

Abstract

This paper describes experiments with speaker verification using support vector machines (SVMs). Verification from coded and uncoded speech is analyzed, both in matched and mismatched conditions. A hybrid SVM-GMM approach is used, in which SVM classifiers with Kullback-Leibler kernel make verification decisions based on the mean values of Gaussian mixtures. The most common narrowband codecs are used, such as G.711, G.729, G.723.1, GSM 06.10, GSM 06.60, and Speex. The Equal Error Rate (EER) is presented for various numbers of Gaussian components, and for various testing conditions. Possible reasons for the non-uniform performance degradation in the case of codec mismatch are discussed. Selected ROC curves are presented. The results are compared with a similar investigation of a close-set speaker classification.

Keywords

speaker recognition; speech coding; support vector machines; EER; G.711; G.723.1; G.729; GSM; Gaussian components; Gaussian mixtures; Kullback-Leibler kernel; SVM classifiers; SVM-GMM approach; SVM-based speaker verification; close-set speaker classification; codec mismatch; coded speech; equal error rate; support vector machines; uncoded speech; Codecs; GSM; Speaker recognition; Speech; Speech coding; Speech recognition; Support vector machines; ROC curve; speaker recognition; speaker verification; speech coding; support vector machine;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Conference (EUSIPCO), 2012 Proceedings of the 20th European

Conference_Location

Bucharest

ISSN

2219-5491

Print_ISBN

978-1-4673-1068-0

Type

conf

Filename

6333978