Synthesis of facial expressions for semantic coding of videophone sequences

Abstract

A method for the synthesis of facial expressions for semantic coding of videophone sequences is presented. Firstly, a generic 3D face model is automatically adapted to the eyes, mouth, eyebrows, nose and chin and cheek contours of the individual face in the sequence. Secondly, a synthesis of facial expressions of this individual face is carried out using a model of the human facial muscles. Applying the described algorithms to the videophone sequences Akiyo and Miss America (CIF 10 Hz), the face model is successfully adapted at the beginning of the videophone sequence and facial expressions of the individual faces are synthesized.