摘要 |
PROBLEM TO BE SOLVED: To obtain a voice recognizing method and a voice recognizing device being not affected by a vocal mode of a user. SOLUTION: A main control section controls a voice synthesizing device 209 and sends voice guidance indicating to describe a word specified by a registered user as an object of voice recognition with Chinese characters and numerals for a registered user, and to send it to a voice service device 203, and in the device 209 reception/registration processing in which a word described with Chinese characters and numerals and sent from a registered user is registered as a registration word responding to this voice guidance is performed. Next, development control processing in which vocal modes of each registration word are developed with HIRAGANA notation (Japanese characters) and controlled is performed. Next, compilation processing in which all vocal modes being able to notate with HIRAGANA are converted to voice information is performed, next, a word for recognizing a voice is obtained from a vocal mode compilation-processed by controlling a voice recognizing device 208, dictionary registration processing in which each registration word is registered in a dictionary for recognizing a voice is performed.
|