摘要 |
<P>PROBLEM TO BE SOLVED: To obtain natural synthesis speech, while simplifying a database. <P>SOLUTION: A speech synthesizer includes: a multiple speaker speech database 17 which is generated by multiple speakers for each sentence or phrase; a text analysis section 3 which receives a text sentence to be synthesized to analyzes the text sentence for each sentence or phrase; and a speech synthesis means for synthesizing speech by using the speech database having a characteristic most similar to the speech database 17, which is used in the sentence or the phrase synthesized just before, in the speech database 17 which has a phoneme coincident with the sentence or the phrase to be synthesized, when the synthetic speech is generated by switching the multiple speaker speech database 17 for each sentence or phrase, which is analyzed by the text analysis section 3. <P>COPYRIGHT: (C)2010,JPO&INPIT |