| 摘要 |
<P>PROBLEM TO BE SOLVED: To excellently recognize a speech that a user utters. <P>SOLUTION: In S31, an utterance decision part 14 stands by until a speech is input from a speech processing part 11 and when the speech is input, an advance to S32 is made to decide whether the detection result of a movement pattern of a specified part from a user's picture is supplied. When it is decided that the detection result is supplied, an advance to S33 is made to decide whether the supplied movement pattern is a pattern of movement during the utterance. When it is decided that the supplied movement pattern is the pattern of movement during the utterance in S33, an advance to S34 is made to report the utterance to a speech recognition part. When receiving the report, the speech recognition part starts recognizing the supplied speech. <P>COPYRIGHT: (C)2007,JPO&INPIT |
