Application of face encoding to art investigations

Through the millennia, one of the primary goals of the art of painting has been an optical depiction of the scenes of the three-dimensional world that form our visual experience. A key role of paintings has been to depict faces in portraiture, conveying the identity of figures of contemporary and historical interest. The effective recognition of such figures requires accurate interpretation of the 3D configuration of the facial features, which is presently a largely subjective art, given the variety of poses employed. A quantitative technique for the analysis of facial expression and 3D facial configuration from the 2D artworks, is applied to the question of how Leonardo da Vinci looked as a young man. There are no accepted portraits of this most dramatic of Renaissance figures as his fame was building, but computational analysis has helped to solve the conundrum of his youthful appearance and unexplained rise to prominence with almost no artistic output.