An Uncertainty Estimation Approach for the Extraction of Source Features in Multisource Recordings - Inria - Institut national de recherche en sciences et technologies du numérique
Conference Papers Year : 2011

An Uncertainty Estimation Approach for the Extraction of Source Features in Multisource Recordings

Abstract

We consider the extraction of individual source features from a multisource audio recording by combining source separation with feature extraction. The main issue is then to estimate and propagate the uncertainty over the separated source signals, so as to robustly estimate the features despite source separation errors. While state-of-the-art techniques were designed for scenarios involving one prominent source plus background noise, we focus on under-determined mixtures involving several sources of interest. We apply either Gibbs sampling or variational Bayes to estimate the posterior probability of the sources and subsequently derive the expectation of the features either by sampling or by moment matching. Experiments over stereo mixtures of three sources show that variational Bayes followed by either feature sampling or moment matching provides the best results for convolutive mixtures, while no improvement is obtained on instantaneous mixtures compared to deterministic feature computation.
Fichier principal
Vignette du fichier
SamplVarGrad.pdf (100.95 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

inria-00597615 , version 1 (06-06-2011)

Identifiers

  • HAL Id : inria-00597615 , version 1

Cite

Kamil Adiloglu, Emmanuel Vincent. An Uncertainty Estimation Approach for the Extraction of Source Features in Multisource Recordings. European Signal Processing Conference (Eusipco 11), Centre Tecnològic de Telecomunicacions de Catalunya, Universitat Politècnica de Catalunya, Aug 2011, Barcelona, Spain. ⟨inria-00597615⟩
193 View
117 Download

Share

More