Automatic adaptation of a face model using action units for semantic coding of videophone sequences | IEEE Journals & Magazine | IEEE Xplore