A Context-Dependent Sequential Decision for Speaker Verification

Hideki NODA
Katsuya HARADA
Eiji KAWAGUCHI

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E82-D    No.10    pp.1433-1436
Publication Date: 1999/10/25
Online ISSN: 
DOI: 
Print ISSN: 0916-8532
Type of Manuscript: LETTER
Category: Speech Processing and Acoustics
Keyword: 
speaker verification,  sequential probability ratio test,  sequential decision,  mean field approximation,  HMM,  

Full Text: PDF>>
Buy this Article



Summary: 
This paper presents an improved method of speaker verification using the sequential probability ratio test (SPRT), which can treat the correlation between successive feature vectors. The hidden Markov model with the mean field approximation enables us to consider the correlation in the SPRT, i. e. , using the mean field of previous state, probability computation can be carried out as if input samples were independent each other.