| 摘要 |
PURPOSE:To ensure a highly efficient operation of an electronic register by voice, by operating the electronic register by voice, confirming the state of operation by a synthetic voice, reciting the state of input and at the same time producing later a synthetic voice of the recital contents. CONSTITUTION:The voice uttered from an operator to a mike 2 is amplified 2 and supplied to a CPU5 via a voice feature extracting circuit 3 and an INPORT4 to be stored in a sampling buffer memory in the CPU5. This voice is then compressed to be stored in a voice pattern register. In this case, if the CPU5 is in the register mode, the input signal is stored in a memory II 12. While in the case of the recognition mode, the contents of the voice pattern register is compared with that of the registered voice pattern of the memory II 12 to select the most approximate contents. Then the utterance of prices, the confirmation of the recognition contents, the request of operation, etc. are produced in a synthetic voice. Furthermore the contents of the supplied price, the sorting, tc. are recited, and the contents of this recital is produced later in the synthetic voice. |
