An Automatic Portrait System Based on an And-or Graph Representation

We have developed an automatic human portrait system based on an And-Or graph representation. The system can automatically generate a set of life-like portraits in different styles from a single frontal face image. The system includes three subsystems, each of which models hair, face, and collar respectively. The face subsystem can be further decomposed into face components: eyebrows, eyes, nose, mouth, and face contour. Each component has a number of distinct sub-templates as a leaf-node in the And-Or graph for portraits. The And-Or graph for portraits is like a "mother template" which produces a large set of valid portrait configurations, each of which is a "composite template" made of a set of sub-templates. We build a large sketch dictionary for portrait components in different styles.  It is thus convenient to change the styles of the portraits by changing the sketch dictionaries.