| 摘要 |
PROBLEM TO BE SOLVED: To provide a music selecting means in which stable recognition precision is obtained inexpensively by practical voice recognition. SOLUTION: The name of a user and the name of a music are inputted from a microphone 3. A voice recognition section 4 recognizes the user's name based on the voice of the inputted user's name and a user's name recognition dictionary 21. A dictionary switching section 5 switches the retrieving object to the registered music dictionary of the user's name from the registered music dictionaries of plural users based on the recognized user's name. A voice recognition section 4 recognizes the name of the music based on the switched and registered music dictionary and the voice of the inputted music name. A music number retrieving section 6 retrieves the music number corresponding to the recognized music name. A display section 8 displays the name of the music of the retrieved music number and a KARAOKE reproducing section 9 reproduces the playing music based on the music data corresponding to the retrieved music number. |
