MMSE estimation of magnitude-squared DFT coefficients with superGaussian priors

@inproceedings{Breithaupt2003MMSEEO,
  title={MMSE estimation of magnitude-squared DFT coefficients with superGaussian priors},
  author={Colin Breithaupt and Rainer Martin},
  booktitle={ICASSP},
  year={2003}
}
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… CONTINUE READING
Highly Cited
This paper has 46 citations. REVIEW CITATIONS

Citations

Publications citing this paper.
Showing 1-10 of 30 extracted citations

Similar Papers

Loading similar papers…