University of Surrey

Test tubes in the lab Research in the ATI Dance Research

Bootstrap Averaging for Model-Based Source Separation in Reverberant Conditions

Chandna, S and Wang, Wenwu (2018) Bootstrap Averaging for Model-Based Source Separation in Reverberant Conditions IEEE/ACM Transactions on Audio Speech and Language Processing.

[img]
Preview
Text
ChandnaWang_TASLP5.pdf - Accepted version Manuscript

Download (412kB) | Preview

Abstract

Recently proposed model-based methods use timefrequency (T-F) masking for source separation, where the T-F masks are derived from various cues described by a frequency domain Gaussian Mixture Model (GMM). These methods work well for separating mixtures recorded in low-to-medium level of reverberation, however, their performance degrades as the level of reverberation is increased. We note that the relatively poor performance of these methods under reverberant conditions can be attributed to the high variance of the frequency-dependent GMM parameter estimates. To address this limitation, a novel bootstrap-based approach is proposed to improve the accuracy of expectation maximization (EM) estimates of a frequencydependent GMM based on an a priori chosen initialization scheme. It is shown how the proposed technique allows us to construct time-frequency masks which lead to improved model-based source separation for reverberant speech mixtures. Experiments and analysis are performed on speech mixtures formed using real room-recorded impulse responses.

Item Type: Article
Divisions : Faculty of Engineering and Physical Sciences > Electronic Engineering
Authors :
NameEmailORCID
Chandna, S
Wang, WenwuW.Wang@surrey.ac.uk
Date : 16 April 2018
Funders : EPSRC
Copyright Disclaimer : © 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
Uncontrolled Keywords : Gaussian mixture model, EM algorithm, bootstrap averaging, model-based source separation, time-frequency masking, reverberant speech mixtures, audio signal processing, spectral histogram
Depositing User : Melanie Hughes
Date Deposited : 16 Jan 2018 12:14
Last Modified : 24 Jan 2018 14:48
URI: http://epubs.surrey.ac.uk/id/eprint/845619

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year


Information about this web site

© The University of Surrey, Guildford, Surrey, GU2 7XH, United Kingdom.
+44 (0)1483 300800