SPEECH SYNTHESIZER, SPEECH SYNTHESIS METHOD FOR LEARNING USER'S BEHAVIOR, AND PROGRAM

摘要

PROBLEM TO BE SOLVED: To synthesize user's favorite speech according to the user, by automatically correcting speech synthesizer based on user's behavior.

SOLUTION: The speech synthesizer includes a speech synthesis section 200 for synthesizing speech; a speech determination section 310, which presents synthesis speech and which outputs user's behavior information for the presented synthesis speech; a user behavior feature extracting section 410 for extracting a user's behavior feature from the user's behavior information; a sound quality evaluation value presumption section 420 for presuming a user's sound quality evaluation value for the synthesis speech from the extracted user's behavior feature; and a correction section 500 for correcting the speech synthesis section 200, based on the sound-quality evaluation value.

COPYRIGHT: (C)2011,JPO&INPIT