Phone adaptive training for speaker diarization

Abstract : The linguistic content of a speech signal is a source of unwanted variation which can degrade speaker diarization performance. This paper presents our latest work to reduce its impact. The new approach, referred to as Phone Adaptive Training (PAT), is analogous to speaker adaptive training used in automatic speech recognition. We report an oracle experiment which shows that PAT has the potential to deliver a 33% relative improvement in the diarization error rate of our baseline system. Practical experiments show significant improvements across two standard, independent evaluation datasets.
Type de document :
Communication dans un congrès
INTERSPEECH 2012, Sep 2012, Portland, U.S. Minor Outlying Islands. pp.1, 2012
Liste complète des métadonnées

Littérature citée [8 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-00733385
Contributeur : Simon Bozonnet <>
Soumis le : mardi 18 septembre 2012 - 15:46:50
Dernière modification le : mardi 18 septembre 2012 - 16:22:56
Document(s) archivé(s) le : mercredi 19 décembre 2012 - 03:45:26

Fichier

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

Identifiants

  • HAL Id : hal-00733385, version 1

Collections

Citation

Simon Bozonnet, Ravichander Vipperla, Nicholas Evans. Phone adaptive training for speaker diarization. INTERSPEECH 2012, Sep 2012, Portland, U.S. Minor Outlying Islands. pp.1, 2012. 〈hal-00733385〉

Partager

Métriques

Consultations de la notice

108

Téléchargements de fichiers

112