Speaker Identication Using Discriminative Learning of Large Margin GMM

Abstract : Gaussian mixture models (GMM) have been widely and suc- cessfully used in speaker recognition during the last decades. They are generally trained using the generative criterion of maximum likelihood estimation. In an earlier work, we proposed an algorithm for discrimi- native training of GMM with diagonal covariances under a large margin criterion. In this paper, we present a new version of this algorithm which has the major advantage of being computationally highly e cient, thus well suited to handle large scale databases. We evaluate our fast algo- rithm in a Symmetrical Factor Analysis compensation scheme. We carry out a full NIST speaker identi cation task using NIST-SRE'2006 data. The results show that our system outperforms the traditional discrimina- tive approach of SVM-GMM supervectors. A 3.5% speaker identi cation rate improvement is achieved.
Liste complète des métadonnées

Littérature citée [14 références]  Voir  Masquer  Télécharger

Contributeur : Khalid Daoudi <>
Soumis le : dimanche 4 décembre 2011 - 16:32:17
Dernière modification le : vendredi 10 janvier 2020 - 21:09:11
Archivage à long terme le : lundi 5 décembre 2016 - 06:54:16


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


  • HAL Id : hal-00647990, version 1


Khalid Daoudi, Reda Jourani, Régine André-Obrecht, Driss Aboutajdine. Speaker Identication Using Discriminative Learning of Large Margin GMM. International Conference on Neural Information Processing (ICONIP), Nov 2011, Shanghai, China. ⟨hal-00647990⟩



Consultations de la notice


Téléchargements de fichiers