摘要 |
<P>PROBLEM TO BE SOLVED: To make a speech recognition device has high performance by improving recognizing performance of speech recognition by improving the precision of speech recognition results by users even when a plurality of users use the speech recognition device. <P>SOLUTION: The speech recognition device has a speech input part 1, a plurality of speech recognition parts 3a to 3n, a recognition result decision part 4 which decides speech recognition results, a user information database 7 which stores user information databases 7a, ..., and a user information selection part 6 which selects user information data. The speech recognition results that the plurality of speech recognition parts 3a to 3n recognize are statistically processed to output the speech recognition results having the highest accuracy among the speech recognition results of the plurality of speech recognition parts as a total recognition result. <P>COPYRIGHT: (C)2005,JPO&NCIPI |