摘要 |
PROBLEM TO BE SOLVED: To generate a lip sink animation image from a real-time speech or a speech having been sound-recorded without depending upon a program. SOLUTION: An animation image generating device comprises a formant information extraction part 1 which extracts formant information featuring vowels by analyzing a speech, and a lip sink animation image generation part 2 which generates the lip sink animation image following up variation of the extracted formant information. The lip sink animation image generation part 2 comprises a vowel identification part 21 which identifies the vowels from the relative positional relation of the extracted low-order formant frequencies, and a lip animation image selective output part 24 which selects and outputs a prepared lip animation image in synchronism with the identified vowels. COPYRIGHT: (C)2003,JPO |