Mouth Shape Sequence Recognition Based on Speech Phoneme Recognition

As part of the effort to realize human face cartoon automatic synthesization, we explored the mouth shape sequence recognition techniques for continuous natural speech. In this paper we introduce such a method based on speech phoneme recognition. The method consists of five major parts, namely, triphone modeling, search space organization, Viterbi searching, speech phoneme recognition, and path tracking back. At the end of this paper, outputs of such a system are illustrated.