MODELLING THE ALTERNATIVE HYPOTHESIS FOR TEXT-DEPENDENT SPEAKER VERIFICATION

Abstract : This paper describes text-dependent speaker verification as a task involving four classes of trials depending on whether the target speaker or an impostor pronounces the expected pass-phrase or not. These four classes are used to reformulate the log-likelihood ratio traditionally used in text-independent speaker verification. Three formulations of the alternative hypothesis are considered, leading to three new expressions of the verification score. Experiments performed on the publicly available RSR2015 database show a significant improvement compared to existing baseline scores. A relative gain up to 61% in term of minimum cost is achieved when considering that the alternative hypothesis is the union of three sub-hypotheses corresponding to the three existing classes of impostures.
Type de document :
Communication dans un congrès
IEEE International Conference on Acoustic Speech and Signal Processing (ICASSP), May 2014, Florence, Italy
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01927571
Contributeur : Anthony Larcher <>
Soumis le : lundi 19 novembre 2018 - 23:23:21
Dernière modification le : jeudi 22 novembre 2018 - 01:10:00

Fichier

Modelling_the_alternative_hypo...
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01927571, version 1

Collections

Citation

Anthony Larcher, Koing Aik Lee, Bin Ma, Haizhou Li. MODELLING THE ALTERNATIVE HYPOTHESIS FOR TEXT-DEPENDENT SPEAKER VERIFICATION. IEEE International Conference on Acoustic Speech and Signal Processing (ICASSP), May 2014, Florence, Italy. 〈hal-01927571〉

Partager

Métriques

Consultations de la notice

7

Téléchargements de fichiers

3