Lip motion synthesis using a context dependent trajectory hidden Markov model

Abstract

Lip synchronisation is essential to make character animation believeable. In this poster we present a novel technique to automatically synthesise lip motion trajectories given some text and speech. Our work distinguishes itself from other work by not using visemes (visual counterparts of phonemes). The lip motion trajectories are directly modelled using a time series stochastic model called ”Trajectory Hidden Markov Model”. Its parameter generation algorithm can produce motion trajectories that are used to drive control points on the lips directly.