UC San Diego inventors have come up with novel computer vision techniques for identification of facial attractiveness. The measure of attractiveness can be tuned to an individual, in contrast to other techniques which base the measure on an average of many people. The invention has been implemented in a working prototype, and a pilot study has already been conducted.
You can read more at this link: http://mplab.ucsd.edu/~jake/fgr08_attractiveness.pdf
personalized attractiveness detection, computer vision