NMF with time-frequency activations to model non-stationary audio events

Abstract : In this article we propose an extension of non-negative matrix factorization (NMF) based on a source/filter underlying model which permits to well decompose audio signals which contain redundant sounds with strong spectral variations. Our algorithm permits notably to use only one atom per musical event (such as a note) for such non-stationary sounds whereas NMF needs several atoms to represent a non-stationary sound. We have tested our algorithm on real audio data and shown how the proposed decomposition provides a smarter representation of audio spectrograms than standard NMF.
Type de document :
Communication dans un congrès
Proc. of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2010, Dallas, Texas, United States. pp.445--448, 2010
Liste complète des métadonnées

Littérature citée [8 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-00945293
Contributeur : Roland Badeau <>
Soumis le : mardi 25 mars 2014 - 08:21:09
Dernière modification le : jeudi 11 janvier 2018 - 06:23:38
Document(s) archivé(s) le : mercredi 25 juin 2014 - 10:43:59

Fichier

Hennequin-ICASSP2010.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00945293, version 1

Citation

Romain Hennequin, Roland Badeau, Bertrand David. NMF with time-frequency activations to model non-stationary audio events. Proc. of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2010, Dallas, Texas, United States. pp.445--448, 2010. 〈hal-00945293〉

Partager

Métriques

Consultations de la notice

111

Téléchargements de fichiers

110