How can I classify Local Binary Patterns histograms using feed forward network?

I am working on a facial expression recognition system and i am using LBP as a texture descriptor. After detecting the face and the regions of interests , computing the LBP histograms and concatenating them, i want to classify test images in 6 class using a freeforward network with a backpropagation network training function based on the histograms. Can someone tell me the most convenient architecture of the network for this application?

Asked:

on 21 Feb 2013

Community Treasure Hunt

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

Start Hunting!