摘要 |
PROBLEM TO BE SOLVED: To provide a voice recognition device which surely specifies an individual who has spoken, and uses a voice recognition dictionary for specific speaker of the specified individual to improve the recognition precision. SOLUTION: When a voice inputted from an input device connected by radio communication is recognized on-line, an (id) buried in the input device and the voice inputted from the input device are received, and the (id) is detected from the received data, and a user is specified on the basis of the (id). A voice recognition dictionary for the specified user is determined to recognize the inputted voice on-line. Thus the voice recognition peculiar to the speaker is used to recognize the voice with high precision. COPYRIGHT: (C)2003,JPO
|