-
Notifications
You must be signed in to change notification settings - Fork 49
/
est_cgmm.m
81 lines (71 loc) · 2.73 KB
/
est_cgmm.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
function [ lambda_v, lambda_y, R_xn, R_n ] = est_cgmm( ffts )
%EST_CGMM is used to estimate the Complex GMM parameters
%and generate the mask for nosie only and noisy t-f bins
% ffts: M*L*(fft_len/2+1), the multi-channel fft matrix
% lambda_v: the mask for noise only t-f bins
% lambda_y: the mask for noisy t-f bins
% Ry, Rv: the spacial covariance matrix of noisy and noise;;
% M*M*F;
[M, T, F ] = size(ffts);
lambda_v = zeros(T, F);
lambda_y =zeros(T, F);
outer=outProdND(ffts); %M*M*T*F
Ry = squeeze(mean(outer, 3));
R_n = zeros([M, M, F]);
Rv = eye(M);
Rv = reshape(Rv, [size(Rv, 1), size(Rv, 2), 1]);
Rv = repmat(Rv, [1, 1, F]);
phi_y = ones(T, F);
phi_v = ones(T, F);
for iter=1:10
for f=1:F
Ry_f = Ry(:, :, f);
Rv_f = Rv(:, :, f);
if rcond(Ry_f) < 0.0001
Ry_f = Ry_f + rand(M)*0.0001;
end
if rcond(Rv_f) < 0.0001
Rv_f = Rv_f + rand(M)*0.0001;
end
invRy_f = inv(Ry_f);
invRv_f = inv(Rv_f);
y_tf = ffts(:, :, f);
y_y_tf = outProdND(y_tf);
sum_y = zeros(M);
sum_v = zeros(M);
acc_n = zeros(M);
e= eye(M)*0.00000;
for t = 1:T
phi_y(t, f) = (1/M)*(trace(y_y_tf(:, :, t)*invRy_f));
phi_v(t, f) = (1/M)*(trace(y_y_tf(:, :, t)*invRv_f));
kernel_y = y_tf(:, t)' * (1/phi_y(t, f))*invRy_f * y_tf(:, t);
kernel_v = y_tf(:, t)' * (1/phi_v(t, f))*invRv_f * y_tf(:, t);
p_y(t, f) = exp(-kernel_y)/(pi*det(phi_y(t, f)*Ry_f));
p_v(t, f) = exp(-kernel_v)/(pi*det(phi_v(t, f)*Rv_f));
lambda_y(t, f) = p_y(t, f) / (p_y(t, f)+p_v(t, f));
lambda_v(t, f) = p_v(t, f) / (p_y(t, f)+p_v(t, f));
sum_y = sum_y + lambda_y(t, f)/phi_y(t, f)*y_y_tf(:, :, t);
sum_v = sum_v + lambda_v(t, f)/phi_v(t, f)*y_y_tf(:, :, t);
acc_n = acc_n + lambda_v(t, f)*y_y_tf(:, :, t); %for eq(4)
end
R_n(:, :, f) = 1/sum(lambda_y(:, f)) * acc_n; %eq(4)
tmp_Ry_f = 1/sum(lambda_y(:, f)) * sum_y;
tmp_Rv_f = 1/sum(lambda_v(:, f)) * sum_v;
[V1, D1] = eig(squeeze(tmp_Ry_f));
[V2, D2] = eig(squeeze(tmp_Rv_f));
entropy1 = -diag(V1, 0)'/sum(diag(V1, 0)) * log(diag(V1, 0)/sum(diag(V1, 0)));
entropy2 = -diag(V2, 0)'/sum(diag(V2, 0)) * log(diag(V2, 0)/sum(diag(V2, 0)));
if entropy1 > entropy2
Ry(:, :, f) = tmp_Rv_f;
Rv(:, :, f) = tmp_Ry_f;
else
Ry(:, :, f) = tmp_Ry_f;
Rv(:, :, f) = tmp_Rv_f;
end
end
Q = sum(sum(lambda_y .* log(p_y+0.001) + lambda_v .* log(p_v+0.001)))
figure(1)
imagesc(real([flipud(lambda_y');flipud(lambda_v')]));
end
R_xn = Ry;
end