摘要 |
<P>PROBLEM TO BE SOLVED: To provide a speech recognition method and a speech recognition device which are capable of automatically switching among speech recognition units on the basis of speech information. <P>SOLUTION: The speech recognition device includes: a speech power detection unit which detects a speech section of a speech inputted from a user via a microphone to calculate a power of the section; a user-microphone distance calculation unit which converts the power of the section to a distance between the user and the microphone; and a speech recognition dictionary switching unit which selects a plurality of speech recognition units as an output destination of a speech signal of the speech on the basis of the distance. <P>COPYRIGHT: (C)2013,JPO&INPIT |