MMSE estimation of magnitude-squared DFT coefficients with superGaussian priors

We present two minimum mean square error (MMSE) frequency domain estimators of the squared magnitude of a clean speech signal that is degraded by additive noise. These estimators are derived under the assumption that the DFT (discrete Fourier transform) coefficients of the clean speech are best modelled by the Gamma probability distribution function (PDF) instead of the common Gaussian PDF. The statistics of the perturbing noise is the Gaussian PDF in one case and the Laplacian PDF in the other. The estimators are used as noise reduction filters in the experimental evaluation. We give a comparison with a previously derived estimator which uses the Gaussian PDF as the PDF for speech and noise coefficients.

[1]  David Malah,et al.  Speech enhancement using a minimum mean-square error log-spectral amplitude estimator , 1984, IEEE Trans. Acoust. Speech Signal Process..

[2]  Richard V. Cox,et al.  A modular approach to speech enhancement with an application to speech coding , 1999, 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258).

[3]  Rainer Martin,et al.  Spectral Subtraction Based on Minimum Statistics , 2001 .

[4]  Ephraim.  Speech enhancement using a minimum mean square error short-time spectral amplitude estimator , 1984 .

[5]  Rainer Martin,et al.  Speech enhancement using MMSE short time spectral estimation with gamma distributed speech priors , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[6]  Steven F. Boll,et al.  Optimal estimators for spectral restoration of noisy speech , 1984, ICASSP.

[7]  Yariv Ephraim,et al.  A Bayesian estimation approach for speech enhancement using hidden Markov models , 1992, IEEE Trans. Signal Process..