BAYESIAN LEARNING FOR TIME-VARYING LINEAR PREDICTION OF SPEECH

Casamitjana, Adria and Sundin, Martin and Ghosh, Prasanta and Chatterjee, Saikat (2015) BAYESIAN LEARNING FOR TIME-VARYING LINEAR PREDICTION OF SPEECH. In: 23rd European Signal Processing Conference (EUSIPCO), AUG 31-SEP 04, 2015, Nice, FRANCE, pp. 325-329.

PDF Eur_Sig_Pro_Con_325_2015.pdf - Published Version Restricted to Registered users only Download (152kB) | Request a copy

Abstract

We develop Bayesian learning algorithms for estimation of time-varying linear prediction (TVLP) coefficients of speech. Estimation of TVLP coefficients is a naturally underdetermined problem. We consider sparsity and subspace based approaches for dealing with the corresponding underdetermined system. Bayesian learning algorithms are developed to achieve better estimation performance. Expectation-maximization (EM) framework is employed to develop the Bayesian learning algorithms where we use a combined prior to model a driving noise (glottal signal) that has both sparse and dense statistical properties. The efficiency of the Bayesian learning algorithms is shown for synthetic signals using spectral distortion measure and formant tracking of real speech signals.

Item Type:	Conference Proceedings
Series.:	European Signal Processing Conference
Publisher:	IEEE
Additional Information:	Copy right for this article belongs to the IEEE, 345 E 47TH ST, NEW YORK, NY 10017 USA
Keywords:	Time-varying linear prediction; sparsity; Bayesian learning; expectation-maximization
Department/Centre:	Division of Electrical Sciences > Electrical Engineering
Date Deposited:	19 Aug 2016 09:40
Last Modified:	19 Aug 2016 09:40
URI:	http://eprints.iisc.ac.in/id/eprint/54298

Actions (login required)

View Item