gss.m

这个MATLAB程序是用于对音频信号进行降噪处理

function H = gss(gamma_1,gamma_2,X)
% GSS - Generalised Spectral Subtraction (Berouti, 1979)
%   H = GSS(GAMMA_1,GAMMA_2,X) returns an array of filter coefficients H
%   corresponding to generalised spectral subtraction given parameters
%   GAMMA_1, and GAMMA_2, and the (possibly square root of) the a posteriori
%   SNR X:
%
%     ( 1 - X^-GAMMA_1 )^GAMMA_2   if X > 1
% H = 
%     0        otherwise
%
% References:
%
% M. Berouti, R. Schwartz, and J. Makhoul.  Enhancement of speech corrupted
% by acoustic noise.  In Proceedings of the IEEE International Conference on
% Acoustics, Speech and Signal Processing, pages 208-211, 1979.
%
% [this form]
% N. Virag.  Single channel speech enhancement based on masking properties
% of the human auditory system.  IEEE Transactions on Speech and Audio
% Processing, 7(2):126-137, March 1999.

% Author: Patrick J. Wolfe
%         Signal Processing Group
%         Cambridge University Engineering Department
%         [e-mail]
%
% Copyright (c) 2000 by Patrick J. Wolfe
% $Revision: 1.2 $  $Date: 2001/01/24 13:07:35 $

% Comments:
% Wiener filter is gss(1,-1,-xihat);

H = zeros(size(X));
ind = find( X > 1 );
if  ~length(gamma_1), gamma_1 = gamma_1(ind); end
if  ~length(gamma_2), gamma_2 = gamma_2(ind); end
H(ind) = ( 1 - X(ind).^-gamma_1 ).^gamma_2;