发明名称 VOICE DATABASE CREATION SYSTEM, VOICE DATABASE CREATION METHOD, AND PROGRAM
摘要 PROBLEM TO BE SOLVED: To provide a voice database (DB) which reduces loads on a user and realizes a high-quality voice close to the quality of the voice of the speaker (user).SOLUTION: The voice quality of the voice input by a specific user reading aloud a first text is evaluated to select an existing voice database having the voice quality closest to that of the specific user, from among plural prepared beforehand existing voice databases having a voice quality different from each other. The voice input by the specific user reading aloud a second text different from the first text is used to create labeled voice fragments of the specific user. The voice fragments of the selected existing voice database are replaced with the labeled voice fragments of the specific user according to a preset fragment replacement rule to update the voice database.
申请公布号 JP2015052748(A) 申请公布日期 2015.03.19
申请号 JP20130186498 申请日期 2013.09.09
申请人 HITACHI ULSI SYSTEMS CO LTD 发明人 SON KEIKA;NAGAMATSU KENJI;MORI TAKEO;KIDO TAKANORI
分类号 G10L13/10;G10L13/06 主分类号 G10L13/10
代理机构 代理人
主权项
地址