| 摘要 |
PURPOSE:To obtain a voice output answer having no sense of incongruity. CONSTITUTION:A voice (the other party name) inputted to a microphone 1 is outputted to a voice recognizing part 7 through an A/D converter 2, subjected to voice recognition, and thereafter, a sound signal (the other party name) corresponding to its recognition result is read out of a user memory 4, and also, an additional voice 'are you Mr.?' stored in a system memory 5 is read out, and the additional voice 'are you Mr.?' is outputted to a pitch converting part 12 continuously from the sound signal (the other party name). In the pitch converting part 12, a pitch of the sound signal (the other party name) outputted from a voice answer control part 3, and a pitch of the additional voice 'are you Mr.?' are converted to the same pitch. Subsequently, in a D/A converter 13, a D/A conversion processing is performed, and thereafter, a voice 'are you Mr.? (the other party name)? ' in which the voice (the other party name)' and the additional voice 'are your Mr.?' are continued is outputted from a loudspeaker 14 as a voice for confirming a result of recognition. |
