| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide voice guidance matching circumstances or user-friendly voice guidance. <P>SOLUTION: In occurrence of an event for which voice guidance matching text data serving as voice guidance data stored in a storage part 10 in any of slave machines 1 should be given, a control part 11 in the slave machine 1 transmits the text data to a master machine 2 via a network 3. Subsequently, in a voice synthesis part 20 of the mater machine 2, an analysis part generates text synthesis data by executing language analysis, which uses information in a language database, on the text data. Then, a voice waveform generation part generates voice signal data produced by automatically switching a voice guidance tone according to content of the voice guidance based on the text synthesis data and tone data from a context dependent HMM. At last, a notification part 21 reproduces the voice signal data to give the voice guidance. <P>COPYRIGHT: (C)2007,JPO&INPIT |
