Hi,,I want to use PAC in face recognition in matlab
I wrote a command for PCA but stucked with writing a code to recover the image from EigenFace
could you please tell me the steps which I need?
this is which I have done so far
load('att.mat');
[n,d] = size(Data);
% calculate the mean
M=mean(Data); % the maxumum value of the mean is 8.9269exp(-09) which means the mean is very close to zero
MM = repmat(M,n,1);
% center the data
x1 = Data-MM;
%compute the covariance matrix
C = cov(x1);
% Compute SVD (equivalently eigen-decomposition) of the convariance matrix
[V, E] = svd©; %V is the eigenvector,E the eigenvalue
plot(E,'ro')
V1 = V(:,1:2); %choose the first 2 column
%transformation Vk? maps a generic data point in R33 to a lower dimensional space R4
y1 = V1'*x1'; %project the data onto the the four components
% calculate the similarity of the input to each training image
Recon-image =V*( V1' * (Data(:,1)-M))+M;
similarity_score = arrayfun(@(n) 1 / (1 + norm(y1(:,n) - Recon-image)), 1:d);
% find the image with the highest similarity
[match_score, match_ix] = max(similarity_score);