| 摘要 |
PROBLEM TO BE SOLVED: To enhance convenience by jointly using at least two modes among a key-word mode, a trigger mode, and a PTT(press-to-talk) mode, as a method of specifying a speech zone for voice recognition. SOLUTION: When a voice recognition function starts, whether a key word is spoken and whether a talk button 1 is operated are found. In the event that a key word is spoken, the function branches into voice recognition processing of key-word mode, and a starting point of a speech zone and an ending point thereof are estimated by a voice-zone detecting part 7. In the event that the talk button 1 is operated, a starting point of a speech zone is determined at this point of time. Whether the button operation is momentary or not is found. If it is momentary, the function branches into voice recognizing processing of trigger mode, and an ending point of the speech zone is estimated by the voice-zone detecting part 7. In the event the operation of the talk button 1 is continuous, the function branches into voice recognizing processing of PTT mode, and an ending point of the speech zone is determined at the point of time the operation of the talk button 1 is released.
|
