| 摘要 |
PURPOSE:To input an original at a high speed by providing a CPU dedicated to input a featured parameter and a CPU dedicated to comparative arithmetic in a voice recognizing device, and executing individually and simultaneously the input of featured parameters and comparative arithmetic. CONSTITUTION:A parameter extracting device 2 extracts the feature parameter FP1 of a voice signal inputted through a microphone 1. The voice recognizing device 3 compares the extracted feature parameter FP1 with a previously stored monosyllable featured parameter FP2A or word featured parameter FP2B to recognize the voice. A KANJI (Chinese character) processor 4 performs KANA (Japanese syllabary)-KANJI conversion of the output code of the voice recognizing device 3 to obtain a KANA and a KANJI code. A mode discriminating device 6A judges whether the voice input VS is recognized as monosyllables or words. Then, whether the recognition result is in cursive form or square form of KANA is set with a KANA code changeover switch 7. |
