DocumentCode
2317303
Title
Using vector quantization in Automatic Speaker Verification
Author
Hayet, D. Jellai ; Tayeb, Laskri Mohamed
Author_Institution
Dept. of Comput. Sci., Badji Mokhtar Univ., Annaba, Algeria
fYear
2012
fDate
24-26 March 2012
Firstpage
1
Lastpage
5
Abstract
This article investigates several technique based on vector quantization (VQ) and maximum a posteriori adaptation (MAP) in Automatic Speaker Verification ASV. We propose to create multiple codebooks of Universal Background Model UBM by Vector Quantization and compare them with traditional approach in VQ, MAP adaptation and Gaussian Mixture Models.
Keywords
maximum likelihood estimation; speaker recognition; ASV; Gaussian mixture models; MAP adaptation; UBM; VQ; automatic speaker verification; maximum a posteriori adaptation; multiple codebooks; universal background model; vector quantization; Adaptation models; Computational modeling; Speech; Training; Vector quantization; Vectors; Automatic Speaker Recognition; False Acceptance; False Rejection; Gaussian Mixture Models; Impostor Models; Linde Buzo Gray Algorithm; Speaker Verification; Vector Quantization;
fLanguage
English
Publisher
ieee
Conference_Titel
Information Technology and e-Services (ICITeS), 2012 International Conference on
Conference_Location
Sousse
Print_ISBN
978-1-4673-1167-0
Type
conf
DOI
10.1109/ICITeS.2012.6216611
Filename
6216611
Link To Document