Fast factorization-based inference for Bayesian harmonic models

Abstract : Harmonic sinusoidal models are a fundamental tool for audio signal analysis. Bayesian harmonic models guarantee a good resynthesis quality and allow joint use of learnt parameter priors and auditory motivated distortion measures. However inference algorithms based on Monte Carlo sampling are rather slow for realistic data. In this paper, we investigate fast inference algorithms based on approximate factorization of the joint posterior into a product of independent distributions on small subsets of parameters. We discuss the conditions under which these approximations hold true and evaluate their performance experimentally. We suggest how they could be used together with Monte Carlo algorithms for a faster sampling-based inference.
Type de document :
Communication dans un congrès
2006 IEEE Int. Workshop on Machine Learning for Signal Processing, Sep 2006, Maynooth, Ireland. pp.117--122, 2006
Liste complète des métadonnées

Littérature citée [10 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/inria-00544652
Contributeur : Emmanuel Vincent <>
Soumis le : mercredi 8 décembre 2010 - 15:58:43
Dernière modification le : mardi 24 octobre 2017 - 17:14:02
Document(s) archivé(s) le : jeudi 10 mars 2011 - 11:29:27

Fichier

vincent_MLSP06.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : inria-00544652, version 1

Citation

Emmanuel Vincent, Mark Plumbley. Fast factorization-based inference for Bayesian harmonic models. 2006 IEEE Int. Workshop on Machine Learning for Signal Processing, Sep 2006, Maynooth, Ireland. pp.117--122, 2006. 〈inria-00544652〉

Partager

Métriques

Consultations de la notice

52

Téléchargements de fichiers

60