Speech enhancement based upon hidden Markov modeling

Author

Ephraim, Yariv ; Malah, David ; Juang, Biing-hwang

Author_Institution

AT&T Bell Lab., Murray Hill, NJ, USA

fYear

1989

fDate

23-26 May 1989

Firstpage

353

Abstract

A maximum a posteriori approach for enhancing speech signals which have been degraded by statistically independent additive noise is proposed. The approach is based upon statistical modeling of the clean speech signal and the noise process using long training sequences from the two processes. Hidden Markov models (HMMs) with mixtures of Gaussian autoregressive (AR) output probability distributions are used to model the clean speech signal. A low-order Gaussian AR model is used for the wideband Gaussian noise considered here. The parameter set of the HMM is estimated using the Baum or the EM (estimation-maximization) algorithm. The enhancement of the noisy speech is done by means of reestimation of the clean speech waveform using the EM algorithm. An approximate improvement of 4.0-6.0 dB in signal-to-noise ratio (SNR) is achieved at 10 dB input SNR

Keywords

Markov processes; speech analysis and processing; Baum algorithm; Gaussian autoregressive model; HMM; SNR; clean speech signal; estimation-maximisation algorithm; hidden Markov models; long training sequences; maximum a posteriori approach; noisy speech; output probability distributions; speech enhancement; statistical modeling; statistically independent additive noise; wideband Gaussian noise; Additive noise; Degradation; Gaussian noise; Hidden Markov models; Probability distribution; Signal processing; Signal to noise ratio; Speech enhancement; Speech processing; Wideband;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1989. ICASSP-89., 1989 International Conference on

Conference_Location

Glasgow

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1989.266438

Filename

266438