摘要 |
PURPOSE:To rapidly sound record the utter voice in a recording and reproducing memory at the time of confirming a recognized result by recording an inputted standard voice in the recording and reproducing memory, and storing the standard pattern in a voice memory and forming the address correspondence between both memories on a table memory. CONSTITUTION:The voice pattern is compared with the standard voice pattern in the voice memory 5 through the pattern formation part 4 of an inputted voice from a microphone 1, and the inputted voice is recognized through a discrimination part 6, etc. Then, the inputted standard voice from the microphone 1 for confirming the recognized result is recorded in the recording and reproducing memory 70, and is confirmed based on the voice pattern and the voice pattern in the memory 5. At this time, the address correspondence between the address of the memory 70 and the address of the memory 5 is formed on the table memory 11 through a control part 10. By such a system recording the voice in the recording and reproducing memory without through a buffer memory, the recording of the utter voice in the recording and reproducing memory at the time of confirming the recognized result is performed rapidly. |