| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide a text speech synthesizer and a method, capable of generating stable and high quality speech synthetic sound. <P>SOLUTION: The text speech synthesizer, which converts a text data to speech, includes: a speech synthetic means for generating a plurality of synthetic speech candidates by different synthetic characteristics on the basis of the same input text; a sound quality evaluation means for evaluating sound quality on each synthetic candidate from the speech synthetic means; and a synthetic speech selection means for selecting one of the synthetic speech candidates as a synthetic speech, on the basis of an evaluation value of each synthetic candidate which is evaluated by the sound quality evaluation means. <P>COPYRIGHT: (C)2010,JPO&INPIT |
