PURPOSE: To realize an audio visual dubbing technique in which the mouth movement of an original speaker can be made to coincide with the movement in a language to be dubbed.
CONSTITUTION: In replacing the original sound track of a video sequence portraying the original speaker with a sound signal indicating the replaced voice of a new speaker, the frame of the sequence stored or transmitted is related to the information of the facial characteristics corresponding to the vocalization in the language to be dubbed. The frame of the sequence is corrected by coinciding the mouth shape of the original speaker with a set of the mouth shape independent of the new speaker's or that of an actor's, by means of the information of the facial characteristics using a lookup table relating to the vocalization of the new speaker to be detected. With an audience of a program selecting a dubbed edition, the mouth movement of the original speaker is adjusted to that of a new speaker's vocalization using a second sound track.
TSUUHAN CHIEN
BAARIN JIEFURII HASUKERU
ERITSUKU DEEBITSUDO PETAJIYAN
