摘要 |
PURPOSE: A method and a system for speech synthesis are provided to improve the voice synthesis quality by adaptively updating the corpus when a user uses a system. CONSTITUTION: A speech synthesis operation for textual contents is performed based on a current acoustical unit set of a corpus(S304), and it is judged whether or not the current acoustical unit for the corpus is updated(S305). The textual contents are analyzed, and the list for target units having a corresponding context feature is generated. Based on the acoustical unit set which is more abundant than the current acoustic unit set for the corpus, the plural acoustic unit candidates are selected fro the each target unit. According to the plural unit candidates, the acoustic units proper to the acoustic synthesis for the textual contents are determined(S306).
|