forum.alglib.net
http://forum.alglib.net/

Classification (face recognition)
http://forum.alglib.net/viewtopic.php?f=2&t=3803
Page 1 of 1

Author:  Ivan12345 [ Sun May 07, 2017 11:23 am ]
Post subject:  Classification (face recognition)

Hi,

I have 12 double values - features of the face. I want to train the neural network to recognize the definite face. (the face of one person)
But the function:
Code:
void mlpcreatetrainercls(int nin, int nclasses, out mlptrainer s)
allows to create at least 2 classes.
1. Is it possible to create the network with one class? How to learn network to recognise the similarity of the faces: (similarity of the input face with the face from the learning set).
I expect network returning the percentage values of similarity (0.7(9) - the high similarity , 0.3(0.5) - the low similarity).
2. If the above task is hardly ever possible, please let me know if it is worth to learn network to recognise several faces? (and to use several classes).
Please advice.

Thanks,
Ivan

Page 1 of 1 All times are UTC
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/