摘要 |
PROBLEM TO BE SOLVED: To provide a speech database and a speech synthesizer that can improve the sound quality of speech synthesis, while maintaining the capacity of speech data, without performing an operation to store the speech data afterwards. SOLUTION: The speech database 100 stores speech data 110, comprising waveform data of a speech and pieces of label information 121, 122, and 123 defining at least one elementary unit, on the basis of the speech data. The label information includes boundary information defining sections of respective elementary units by dividing the speech data, and phoneme information defining the phoneme names of the respective elementary pieces, and a plurality of pieces of label information corresponding to the speech data. COPYRIGHT: (C)2008,JPO&INPIT
|