“Speech to talking heads system based on hidden Markov models” by Yotsukura, Morishima and Nakamura

  • ©Tatsuo Yotsukura, Shigeo Morishima, and Satoshi Nakamura




    Speech to talking heads system based on hidden Markov models



    This paper describes a technique to create human-like talking head speech animation with levels of naturalness and realism by mapping from speech information to facial movement sequences. Speech animation techniques for human-like natural talking head systems have traditionally included both key-framing methods [Cohen and Massaro 1993]and physics-based methods[Waters 1987]. Recent machine learning methods provide a new technique for speech animation systems


    1. Cohen, M. M., Massaro, D. W. 1993. Modeling coarticulation in synthetic visual speech. Models and Techniques in Computer Animation, pp. 139–156.
    2. Yamamoto, E., Nakamura, S., Shikano, K. 1998. Lip movement synthesis from speech based on hidden markov models. Speech Communication, pp. 105–115
    3. Waters, K. 1987. A muscle model for animating three-dimensional facial expressions. ACM SIGGRAPH 87, pp. 17–24.

ACM Digital Library Publication:

Overview Page: