Computational Methods For Structured Sparse Component Analysis of Convolutive Speech Mixtures

Abstract : We cast the under-determined convolutive speech separation as sparse approximation of the spatial spectra of the mixing sources. In this framework we compare and contrast the major practical algorithms for structured sparse recovery of speech signal. Specific attention is paid to characterization of the measurement matrix. We first propose how it can be identified using the Image model of multi-path effect where the acoustic parameters are estimated by localizing a speaker and its images in a free space model. We further study the circumstances in which the coherence of the projections induced by microphone array design tend to affect the recovery performance.
Type de document :
Communication dans un congrès
Proceeding of the 37th International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Mar 2012, tokyo, Japan. 2012
Liste complète des métadonnées

https://hal.inria.fr/hal-00700391
Contributeur : Jules Espiau de Lamaestre <>
Soumis le : mardi 22 mai 2012 - 17:55:30
Dernière modification le : jeudi 26 octobre 2017 - 16:34:02

Identifiants

  • HAL Id : hal-00700391, version 1

Collections

Citation

Afsaneh Asaei, Mike Davies, Hervé Bourlard, Volkan Cevher. Computational Methods For Structured Sparse Component Analysis of Convolutive Speech Mixtures. Proceeding of the 37th International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Mar 2012, tokyo, Japan. 2012. 〈hal-00700391〉

Partager

Métriques

Consultations de la notice

94