On comparing and combining intra-speaker variability compensation and unsupervised model adaptation in speaker verification

2007 ◽  
Author(s):  
Claudio Garreton ◽  
Nestor Becerra Yoma ◽  
Fernando Huenupán ◽  
Carlos Molina
Keyword(s):  
Speaker Verification ◽  
Model Adaptation ◽  
Speaker Variability ◽  
Variability Compensation

Intra-speaker variability compensation in speaker verification with limited enrolling data

2006 ◽  
Author(s):  
Claudio Garreton ◽  
Nestor Becerra Yoma ◽  
Carlos Molina ◽  
Fernando Huenupan
Keyword(s):  
Speaker Verification ◽  
Speaker Variability ◽  
Variability Compensation

Confidence measure based unsupervised target model adaptation for speaker verification

2007 ◽  
Author(s):  
A. Preti ◽  
Jean-François Bonastre ◽  
Driss Matrouf ◽  
F. Capman ◽  
B. Ravera
Keyword(s):  
Speaker Verification ◽  
Model Adaptation ◽  
Confidence Measure ◽  
Target Model

Improving speaker verification performance against long-term speaker variability

2016 ◽  
Vol 79 ◽  
pp. 14-29 ◽  
Author(s):  
Linlin Wang ◽  
Jun Wang ◽  
Lantian Li ◽  
Thomas Fang Zheng ◽  
Frank K. Soong
Keyword(s):  
Speaker Verification ◽  
Speaker Variability

Speaker Verification with Fuzzy Fusion and Genetic Optimization

1999 ◽  
Vol 3 (6) ◽  
pp. 451-456
Author(s):  
Tuan Pham ◽  
◽  
Michael Wagner ◽  
Keyword(s):  
Genetic Algorithms ◽  
Speaker Verification ◽  
A Priori ◽  
Similarity Measures ◽  
Error Rates ◽  
Fuzzy Integral ◽  
Speech Corpus ◽  
Speaker Variability ◽  
Verification System ◽  
Fuzzy Fusion

Most speaker verification systems are based on similarity or likelihood normalization techniques as they help to better cope with speaker variability. In the conventional normalization, the it a priori probabilities of the cohort speakers are assumed to be equal. From this standpoint, we apply the fuzzy integral and genetic algorithms to combine the likelihood values of the cohort speakers in which the assumption of equal <I>a priori</I> probabilities is relaxed. This approach replaces the conventional normalization term by the fuzzy integral which acts as a non-linear fusion of the similarity measures of an utterance assigned to the cohort speakers. Furthermore, genetic algorithms are applied to find optimal fuzzy densities which are very important for the fuzzy fusion. We illustrate the performance of the proposed approach by testing the speaker verification system with both the conventional and the proposed algorithms using the commercial speech corpus TI46. The results in terms of the equal error rates show that the speaker verification system using the fuzzy integral is more favorable than the conventional normalization method.

Sign in / Sign up

Export Citation Format

Share Document