| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To provide a talking unit which collectively selects voices of previously registered voiceprints for identifying collected signals by a microphone but will not select unregistered sounds e.g. the noise around the calling device, sounds of television sets, etc. <P>SOLUTION: Gain-variable amplifiers 301-306 are provided as means for amplifying detected signals by microphones MC1-MC6. When a DSP 25 selects e.g. a high-pitched collected sound signal, a voiceprint authenticator 32 authenticates whether the selected sound signal agrees with a registered voiceprint. If agreeing with it, an amplifier gain adjuster 34 raises the gain of the variable-gain amplifier corresponding to one microphone selected by the DSP 25 to finally output the voiceprint-authenticated sound signal of the microphone out of a LINE OUT speaker 16, etc. <P>COPYRIGHT: (C)2005,JPO&NCIPI</p> |
