Audiovisual-to-articulatory speech inversion using Active Appearance Models for the face and Hidden Markov Models for the dynamics.

Athanassios Katsamanis, George Papandreou, Petros Maragos

Browse the full ICASSP paper archive.