Emotional speaker verification with linear adaptation

Author

Fanhu Bie ; Dong Wang ; Zheng, Thomas Fang ; Ruxin Chen

Author_Institution

Dept. of Comput. Sci. & Technol., Tsinghua Univ., Beijing, China

fYear

2013

fDate

6-10 July 2013

Firstpage

91

Lastpage

94

Abstract

Speaker verification suffers from significant performance degradation on emotional speech. We present an adaptation approach based on maximum likelihood linear regression (MLLR) and its feature-space variant, CMLLR. Our preliminary experiments demonstrate that this approach leads to considerable performance improvement, particularly with CMLLR (about 10% relative EER reduction in average). We also find that the performance gain can be significantly increased with a large set of training data for the transform estimation.

Keywords

emotion recognition; maximum likelihood estimation; regression analysis; speaker recognition; CMLLR; emotional speaker verification; feature-space variant; linear adaptation approach; maximum likelihood linear regression; relative EER reduction; training data; transform estimation; Acoustics; Adaptation models; Hidden Markov models; Speech; Training; Transforms; Vectors; MLLR; emotional speech; speaker verification;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal and Information Processing (ChinaSIP), 2013 IEEE China Summit & International Conference on

Conference_Location

Beijing

Type

conf

DOI

10.1109/ChinaSIP.2013.6625304

Filename

6625304