Markov chain Monte Carlo methods for speech enhancement

Author

Vermaak, Jaco ; Niranjan, Mahesan

Author_Institution

Dept. of Eng., Cambridge Univ., UK

Volume

2

fYear

1998

fDate

12-15 May 1998

Firstpage

1013

Abstract

This paper investigates a Bayesian approach to the enhancement of speech signals corrupted by additive white Gaussian noise. Parametric models for the speech and noise processes are constructed, leading to a posterior distribution for the model parameters and uncorrupted speech samples given the observed noisy speech samples. Being analytically intractable, inferences concerning these variables are performed using Markov chain Monte Carlo (MCMC) methods. The efficiency of the sampling scheme within this framework is further improved by employing state-space techniques based on the Kalman filter

Keywords

Bayes methods; Gaussian noise; Kalman filters; Markov processes; Monte Carlo methods; filtering theory; signal sampling; speech enhancement; state-space methods; white noise; Bayesian approach; Kalman filter; Markov chain Monte Carlo methods; a posterior distribution; additive white Gaussian noise; observed noisy speech samples; parametric models; sampling scheme; speech enhancement; state-space techniques; uncorrupted speech samples; Additive noise; Additive white noise; Bayesian methods; Electronic mail; Gaussian noise; Monte Carlo methods; Parametric statistics; Sampling methods; Speech enhancement; Speech processing;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on

Conference_Location

Seattle, WA

ISSN

1520-6149

Print_ISBN

0-7803-4428-6

Type

conf

DOI

10.1109/ICASSP.1998.675439

Filename

675439