| 摘要 |
PROBLEM TO BE SOLVED: To provide a speech recognition device and program capable of easily confirming a speech while looking at text.SOLUTION: Speech data is divided to a prescribed length and turned into text. Each of divided speech data is searched for information about a position (an elapsed time from the beginning or an address, etc.) in the original speech data and correlated with the text data. The text data turned into text is displayed on a display unit 25. A user designates, via a user interface 26, a portion of each text data in which the user wants to confirm speech. A CPU 22 reads out position information correlated with the designated text data and generates speech data from the position indicated in the position information. |
