PURPOSE: To realize clear talking by automatically adjusting at least one of a voice signal inputted from a microphone and a conversion voice signal outputted from a speaker based on a noise data detected by a voice recognition device at hand-free talking.
CONSTITUTION: A voice recognition device 8 uses a voice characteristic extraction circuit to extract a noise data for a prescribed period at hand-free talking and sends the result to a noise discrimination circuit 11. The circuit 11 implements time average processing and change extraction processing of a spectrum characteristic of noise and discriminates the noise data from the signal obtained through the processing. Then the circuit 11 discriminates it that a change in the characteristic is steady in the noise spectrum and the quantity of the noise level to adjust the gain of a gain variable amplifier 13. On the other hand, the gain of the gain variable amplifier 12 is adjusted even to the microphone 4 to allow the voice signal to be easily heard by an opposite party.
JPH06225037 | FACSIMILE EQUIPMENT |
JP3001493 | [Name of device] Emergency call device |
JP2000134306 | AUTOMATIC DIALING DEVICE |
SHITAYA MITSUO
KANEKO KAZUMA
Next Patent: STATION SUBSCRIBER DATA GENERATING METHOD FOR PRIVATE BRANCH OF EXCHANGE