Image Categorizing

This program uses my SVM implementation to figure out the "ideal" type of an image. You give it a bunch of positive and negative samples, and it figures out an image that correlates the most with the positives and the least with the negatives.

The cool part of this program is that it saves the image it finds to a file. In sample_data, I have demonstrated how it can figure out the "average" pair of eyes, highlighting key features of eyeballs, a nose, etc.