SOUND-ASSISTED IMAGE PROCESSING

摘要

PROBLEM TO BE SOLVED: To provide a method and means for acoustically supported picture processing by which videos are synthesized into animations by using parameters extracted from audio signals. SOLUTION: After audio signals are sampled and a first beseem sequence is generated at the first rate corresponding to the sampling rate in the audio area of the sampled audio signals in response to the sampled audio signals (120) and the first beseem sequence is converted into a second beseem sequence by using a conversion standard at the second rate corresponding to the frame rate in a video area (130), a picture is processed in response to the second beseem sequence. For example, the structure of the face model of a speaker is deformed so that the mouth section of the video of the face of the speaker can move in accordance with voices in response to a rate-converted beseem sequence extracted from audio signals while the video of the face of the speaker is used as an animation by using a three-dimensional wire-frame face model which maps the surface texture of the speaker.