How do I compare the face data in Linear Discriminant Analysis face recognition

1 view (last 30 days)
Well, I have two questions.
I have a dataset containing 80 images divided into 16 classes, i.e each class contains 5 images. Each image is of 200*180px. So my dataset has the dimensions 36000x80.
First, According to the algorithms, my training vectors should have the dimensionality as M-1, i.e. 15, in this case. Also since there are 16 classes, should the vectors also be 15?
I mean, should the training vector matrix be 15x15?
According to the algorithms, the matrix comes out to be 15x80, i.e the number of images rather than the classes.
Which one is correct, 15x15 or 15x80?
And second...
If 15x80 is correct, how do I compare the data? I mean, compare the test vector with 80 vectors?
Shouldnt it be just comparing the class-wise information, i.e. just 15 vectors?

Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!