Speaker identification based on discriminative vector quantization

Author

Zhou, Guangyu ; Mikhael, Wasfy B.

Author_Institution

Dept. of Electr. & Comput. Eng., Central Florida Univ., Orlando, FL

Volume

2

fYear

2003

fDate

30-30 Dec. 2003

Firstpage

617

Abstract

A novel discriminative vector quantization method for speaker identification (DVQSI) is proposed, and its parameters selection is discussed. The vector space of speech features is divided into a number of subspaces and the distribution of the inter speaker variation inside the speech feature vector space is considered. Discriminative weighted average distortion instead of equally weighted average distortion is used in speaker identification (SI). The proposed approach can be considered a generalization of the existing vector quantization (VQ) technique and the experimental results confirm the improved SI accuracy

Keywords

speaker recognition; vector quantisation; discriminative vector quantization; discriminative weighted average distortion; parameter selection; speaker identification; speech features; subspaces; vector space; Distortion measurement; Extraterrestrial measurements; Neural networks; Performance evaluation; Robustness; Sociotechnical systems; Speaker recognition; Speech; Testing; Vector quantization; Discriminative weight; Speaker identification; Subspaces; Vector quantization;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 2003 IEEE 46th Midwest Symposium on

Conference_Location

Cairo

ISSN

1548-3746

Print_ISBN

0-7803-8294-3

Type

conf

DOI

10.1109/MWSCAS.2003.1562362

Filename

1562362