Audiovisual-to-articulatory speech inversion using Active Appearance Models for the face and Hidden Markov Models for the dynamics | IEEE Conference Publication | IEEE Xplore