Judgments of Personality Traits from Real World Face Images

Existing models of facial personality trait judgments control a large number of likely sources of variance for real-world trait judgments. ! Our data set includes reliable trait ratings on images with these sources of variance included. ! It is possible to train a computer vision algorithm on trait-rated real-world images that approximates human trait judgments. ! This points to future avenues for empirically studying underexplored factors that contribute to personality judgments. Conclusions Question and Approach