| 摘要 |
<p>A system for generating an animated image of at least a head of a speaking character using a data stream representing text that is to be spoken by the animated character. The system has a processing module for processing the data stream to provide an audio signal representing the text to be spoken and a phoneme signal representing a sequence of phonemes corresponding to successive portions of the audio signal. The output of the processing module is sent to a phoneme to morph module, which is operable to produce a series of morph targets, each morph target defining a pre-determined mouth shape that matches the corresponding phoneme. Also included is a morph randomiser for randomly or pseudo randomly selecting at least one additional morph target and including that in the series of morph targets and an animation randomiser for randomly or pseudo randomly generating animation instructions that define gross head and/or body movement of the character. A generator module uses the animation instructions and the morph targets to generate an animated image, which image is synchronised with the audio signal.</p> |
