Personalized facial attractiveness prediction

13 years 11 months ago

Download mplab.ucsd.edu

We present an approach to learning the personal preferences of individual users directly from example images. The target application is computer assisted search of partners in online dating services. The proposed approach is based on the use of ε-SVMs to learn a regression function that maps low level image features into attractiveness ratings. We present empirical results based on a dataset of images collected from a large online dating site. Our system achieved correlations of up to 44 % (Pearson rank correlation) on the attractiveness predictions for individual users. We show evidence that the approach learned not just on a universal sense of attraction shared by multiple users, but capitalized on the preferences of individual subjects. Our results are promising and could already be potentially used to facilitate the personalized search of partners in online dating.

Jacob Whitehill, Javier R. Movellan

Real-time Traffic