| 摘要 |
PROBLEM TO BE SOLVED: To provide a voice recognizing device capable of selecting a recognition object word with a simple operation regardless of coincidence with voice when the recognition object word liable to be erroneously recognized exists among recognition candidates. SOLUTION: This voice recognizing device is provided with a microphone 101 for receiving voice, an external storage device 1035 storing a plurality of recognition object words (Kanagawa prefecture 801, etc.), and storing the correspondence between an individual recognition object word and a recognition object word (Nagano prefecture 803) liable to be erroneously recognized, a signal processor 1034 for operating the coincidence of inputted voice with each of the recognition object words and defining a plurality of recognition object words as a recognition candidate on the basis of the coincidence, and an input device 104 that can perform an operation in a first direction 804 and an operation in a second direction 802. A selection is made from the recognition candidates by the operation in the first direction, while a selection is made between the selected recognition candidates and the recognition object words liable to be erroneously recognized by the operation in the second direction. COPYRIGHT: (C)2003,JPO
|
