Abstract

In MPEG-4, Facial Definition Parameters (FDPs) and Facial Animation Parameters (FAPs) are defined to animate 1 a facial object. Most of the previous facial animation reconstruction systems were focused on synthesizing animation from manually or automatically generated FAPs but not the FAPs extracted from natural video scene. In this paper, an analysis-synthesis MPEG-4 visual communication system is established, in which facial animation is reconstructed from FAPs extracted from natural video scene.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call