Speaker Diarization With Unsupervised Training Framework - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

Speaker Diarization With Unsupervised Training Framework

Résumé

This paper investigates single and cross-show diarization based on an unsupervised i-vector framework, on French TV and Radio corpora. This framework uses speaker clustering as a way to automatically select data from unlabeled corpora to train i-vector PLDA models. Performances between supervised and unsupervised models are compared. The experimental results on two distinct test corpora (one TV, one Radio) show that unsupervised models perform as good as supervised models for both tasks. Such results indicate that performing an effective cross-show diarization on new language or new domain data in the future should not depend on the availability of manually annotated data.
Fichier principal
Vignette du fichier
speaker-diarization-unsupervised.pdf (367.99 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01433167 , version 1 (22-03-2017)

Identifiants

Citer

Gaël Le Lan, Sylvain Meignier, Delphine Charlet, Paul Deléglise. Speaker Diarization With Unsupervised Training Framework. 41st IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2016), Mar 2016, Shanghai, China. pp.5, ⟨10.1109/ICASSP.2016.7472741⟩. ⟨hal-01433167⟩
167 Consultations
887 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More