摘要 |
PROBLEM TO BE SOLVED: To provide a device which becomes ready in a short time period and generates a lip-sync animation which realizes a natural motion which effectively reflects a motion of an actual face. SOLUTION: An animation-data generating device includes; a three-set visual element data column generation part 80 which generates phoneme data columns from voice data 42 and converts phoneme labels to visual element labels, by using corresponding relations between the phoneme labels and the visual element labels, to generate three-set visual element data columns; a three-set visual element unit selection part 82 which selects units, whose three-set visual element labels accord with the respective three-set visual element data and which are evaluated as optimal, based on continuation lengths and an average power of voices, from visual element corpora 62; and a three-set visual element unit connection part 84 which connects the face-motion data included in the selected units on a time axis, according to a time sequence of the three-set visual element data, to generate face-animation data. COPYRIGHT: (C)2008,JPO&INPIT
|