A Source Localization/Separation/Respatialization System Based on Unsupervised Classification of Interaural Cues

Abstract : In this paper we propose a complete computational system for Auditory Scene Analysis. This time-frequency system localizes, separates, and spatializes an arbitrary number of audio sources given only binaural signals. The localization is based on recent research frameworks, where interaural level and time differences are combined to derive a confident direction of arrival (azimuth) at each frequency bin. Here, the power-weighted histogram constructed in the azimuth space is modeled as a Gaussian Mixture Model, whose parameter structure is revealed through a weighted Expectation Maximization. Afterwards, a bank of Gaussian spatial filters is configured automatically to extract the sources with significant energy accordingly to a posterior probability. In this frequency-domain framework, we also inverse a geometrical and physical head model to derive an algorithm that simulates a source as originating from any azimuth angle.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-00307889
Contributor : Joan Mouba <>
Submitted on : Tuesday, July 29, 2008 - 3:10:44 PM
Last modification on : Thursday, January 11, 2018 - 6:20:16 AM
Long-term archiving on : Friday, October 5, 2012 - 11:28:09 AM

File

dafx0_p_.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00307889, version 1

Collections

Citation

Joan Mouba, Sylvain Marchand. A Source Localization/Separation/Respatialization System Based on Unsupervised Classification of Interaural Cues. Proceedings of the Digital Audio Effects (DAFx06) Conference, Sep 2006, Canada. pp.233--238. ⟨hal-00307889⟩

Share

Metrics

Record views

417

Files downloads

130