摘要 |
<P>PROBLEM TO BE SOLVED: To provide a speech recognition device, a robot, a speech recognition method, a program and a recording medium, capable of improving accuracy of speech recognition. Ž<P>SOLUTION: The speech recognition device includes: a speech input means for input of speech; a sound source direction detecting means which detects a coming direction of the input speech, and which outputs a sound source direction detecting result; a speech recognition means which detects an utterance section of the input speech, and which outputs a speech recognition result by recognizing speech in the utterance section; a speech detection state obtaining means for output of information of whether the speech recognition means is detecting speech; an output decision means which outputs the sound source direction detecting result output from the sound source direction detection means, when speech is not being detected, and which suppresses output of the sound source, until the time speech is no more detected, when speech is being detected; and an operation means which performs operation according to at least either the sound source direction detecting result output from the output decision means, or the speech recognition result output from the speech recognition result. Ž<P>COPYRIGHT: (C)2010,JPO&INPIT Ž
|