发明名称 METHOD FOR SYNTHESIZING VOICE
摘要 PURPOSE:To obtain a synthesized sound of high quality. CONSTITUTION:A pitch pattern from a pitch pattern data base 6 is normalized by a pitch pattern normalized part 21 and applied to a pitch pattern forming part 5. In the normalizing part 21, a value obtained by subtracting the output of an average pitch part 23 from the output of an analytical pitch part 22 is multiplied by a division output value between the output of a reference pitch maximum change width detecting part 26 and a pitch maximum change detecting part 27 and the output value of a reference average pitch part 30 is added to the multiplied result.
申请公布号 JPH05232991(A) 申请公布日期 1993.09.10
申请号 JP19920033791 申请日期 1992.02.21
申请人 MEIDENSHA CORP 发明人 SAWADA YOSHIMASA
分类号 G10L13/00;G10L21/04;(IPC1-7):G10L3/02;G10L3/00 主分类号 G10L13/00
代理机构 代理人
主权项
地址