| 摘要 |
PURPOSE:To improve the processing ability of the speech input device by providing a discrete input mode in which words are recognized, one by one, and a fast input mode in which they are recognized continuously. CONSTITUTION:A speech recognition device in initial input operation operates while controlled in fast input mode in which the input speed is high. When an operator detects a wrong input, a voice is inputted again and the input mode is switched to the discrete input mode automatically and words are recognized, one by one, to lower the probability of the wrong input, securing the input operation. When the operator detects the wrong input and pushes a wrong input report button, an output gate 3b for re-input request display information 3a is opened and a re-input request is displayed on a display unit 2. Simultaneously, an output gate 3d for a mode switching control signal 3c is opened to send a control signal to the speech recognition unit 1. |
