I have been able to from a pile of used to training image eigenvalue and eigenvector, also can show characteristics of face,
But then how do you how to identify part has little concept of the

Refers to the need to identify the images feature extraction, and then the feature vector and feature vector obtained from the training before doing bad, the relatively small difference is the same person, big difference, it is not the same person, is that right?

CodePudding user response:

Can only feature extraction was carried out on the picture?