摘要 |
PROBLEM TO BE SOLVED: To provide a "speech input device" capable of performing a variety of types of editing using speech for a text which is speech-recognized.SOLUTION: At the start of speech recognition processing of a speech recognition engine 2, a text input mode is set (702). In the text input mode, a recognition result text, which has recognized the speech uttered by a user, is written in text data (703-708), and the mode is switched to a command input mode (711). In the command input mode, an editing operation of the text data is performed according to an editing operation command corresponding to the recognition result text which has recognized the speech uttered by the user (712-728). Also, in the command input mode, if a text in which the user instructs a mode switch is acquired as the recognition result text (731-732), the mode is returned to the text input mode (741). |