Prev: retreiving input from the text box
Next: how to separate this type of data?
From: Nirvana on 26 Apr 2010 02:18
Hi, i am preparing a training set for svm.

My face training image set
300 training image (100 person, 3 image per person). All image reshape
to Nx1 vector.

I am not sure what it mean by
C1 = within class differences set (class +1)
and
C2 = between class differences set. (class -1)

This is what i do,

C1 = [ P1 P2 ... P100 ] *(Nx100) matrix size
'P1 means = person 1 mean face'

C2 = [ X1 X2 ... X100 ] *(Nx100) matrix size
'X1 means = P1 - training_mean_image'

T = [C1 C2]
T is a new training set to be feed for svm.
T matrix size is (Nx200) *N is my image dimension.

Am i doing the right way?
 | 
Pages: 1
Prev: retreiving input from the text box
Next: how to separate this type of data?