Using a Visual Voice Activity Detector to Regularize the Permutations in Blind Separation of Convolutive Speech Mixtures - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2007

Using a Visual Voice Activity Detector to Regularize the Permutations in Blind Separation of Convolutive Speech Mixtures

Résumé

Audio-visual speech source separation consists in mixing visual speech processing techniques (e.g. lip parameters tracking) with source separation methods to improve and/or simplify the extraction of a speech signal fromamixture of acoustic signals. In this paper, we present a new approach to this problem: visual information is used here as a voice activity detector (VAD). Results show that, in the difficult case of realistic convolutive mixtures, the classic problem of the permutation of the output frequency channels can be solved using the visual information with a simpler processing than when using only audio information.
Fichier principal
Vignette du fichier
Rivet_DSP07.pdf (703.32 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte
Loading...

Dates et versions

hal-00173341 , version 1 (19-09-2007)

Identifiants

Citer

Bertrand Rivet, Laurent Girin, Christine Serviere, Dinh-Tuan Pham, Christian Jutten. Using a Visual Voice Activity Detector to Regularize the Permutations in Blind Separation of Convolutive Speech Mixtures. DSP 2007 - 15th IEEE International Conference on Digital Signal Processing, Jul 2007, Cardiff, United Kingdom. pp.223-226, ⟨10.1109/ICDSP.2007.4288559⟩. ⟨hal-00173341⟩
174 Consultations
157 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More