摘要 |
PURPOSE:To improve service with the recognition rate of a speech recognition means bettered by prohibiting the speech recognition treatment of voice input from a microphone when person existence is not detected with a person detecting means or voice output is made with a voice output means. CONSTITUTION:When a user stands in a person detection region, a person detected signal 6a is outputted to a speech recognition device 9 with a person detecting device 6, a call register voice message or a register-unable voice message is outputted with a speaker 8 by a voice output device 11, and a recognition stop signal 11b is outputted to the speech recognition device 9 during the above operation. When the person detected signal 6a is received and the recognition stop signal 11b is not received, if a voice input signal 5a from a microphone 5 is inputted, the voice input signal 5a is compared with an already registered voice with the speech recognition device 9, and when the voices concur, a voice call register signal 9a corresponding to the inputted voice input signal 5a is outputted to a call register device 10. |