A Shift Tolerant Dictionary Training Method

Abstract : Traditional dictionary learning method work by vectorizing long signals, and training on the frames of the data, thereby restricting the learning to time-localized atoms. We study a shift-tolerant approach to learning dictionaries, whereby the features are learned by training on shifted versions of the signal of interest. We propose an optimized Subspace Clustering learning method to accommodate the larger training set for shift-tolerant training. We illustrate up to 50% improvement in sparsity on training data for the Subspace Clustering method, and the KSVD method [1] with only a few integer shifts. We demonstrate improvement in sparsity for data outside the training set, and show that the improved sparsity translates into improved source separation of instantaneous audio mixtures.
Type de document :
Communication dans un congrès
Rémi Gribonval. SPARS'09 - Signal Processing with Adaptive Sparse Structured Representations, Apr 2009, Saint Malo, France. 2009
Liste complète des métadonnées

Littérature citée [7 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/inria-00369548
Contributeur : Ist Rennes <>
Soumis le : vendredi 20 mars 2009 - 11:48:33
Dernière modification le : lundi 20 juin 2016 - 14:10:32
Document(s) archivé(s) le : vendredi 12 octobre 2012 - 14:00:44

Fichier

62.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : inria-00369548, version 1

Collections

Citation

B. Vikrham Gowreesunker, Ahmed H. Tewfik. A Shift Tolerant Dictionary Training Method. Rémi Gribonval. SPARS'09 - Signal Processing with Adaptive Sparse Structured Representations, Apr 2009, Saint Malo, France. 2009. 〈inria-00369548〉

Partager

Métriques

Consultations de la notice

99

Téléchargements de fichiers

108