matlab - What does eigenvalues represent in Face Recognition by Eigenfaces -
i've got set of training face images (40 images). each image size 28*34. there, i'm able eigenvector, score, latent using princomp
function in matlab.
i've got 952 latents (eigenvalues in covariance matrix) in descending form : 4.2785 0 . eigenvalues zeros k=40 onwards.
may know the eigenvalues indicate ? (say bigger value means more significant variance?) how identify best k value (principal component)?
thank !
since have 40 input faces cannot expect have more 40 principal components. therefore eigenvalue becomes 0 k=40 onwards.
visualize results, take 40 leading eigen vectors, reshape
them 28-by-34 , imagesc
them. have got?
Comments
Post a Comment