| 摘要 |
PROBLEM TO BE SOLVED: To provide a speech recognition device that is lowered in false recognition rate.SOLUTION: A speech recognition device includes a storage part which stores a plurality of registered words, a first speech recognition part which compares speech data with the plurality of registered words, a second speech recognition part which makes a determination based upon a comparison result as a result of the comparison, and a control signal generation part. The first speech recognition part presents a first registered word having maximum likelihood among the plurality of registered words as the comparison result to the second speech recognition part, which determines that the first registered word is recognized when the first registered word is presented (n) times (n: an integer equal to or larger than 2) within a predetermined time. The control signal generation part generates a control signal corresponding to the first registered word when it is determined that the first registered word is recognized. |
