| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide a speech recognition system which can improve recognition precision and simplify the operation up to acquisition of a correct recognition result. <P>SOLUTION: An on-vehicle device 100 includes a speech storage buffer 112 which saves a speech that a speaker voices; and a speech recognition part 114 which performs speech recognition processing on the speech saved in the speech storage buffer 112 using a recognition dictionary 116. A facility retrieval server 150 includes a speech recognition part 160 which performs speech recognition processing on the speech saved in the speech storage buffer 112 using a recognition dictionary 162 different from the recognition dictionary 116. A priority adjustment part 122 provided to the on-vehicle device 100 determines a recognition candidate corresponding to the speech saved in the speech storage buffer 112 based upon recognition results of the two speech recognition parts 114, 160. <P>COPYRIGHT: (C)2012,JPO&INPIT |
