PURPOSE: To recognize a sound other than human words and gives information on it.

CONSTITUTION: The speech recognition device is equipped with a speech input means 1, a feature extracting means 2 which extracts features of a speech signal, a feature standard pattern storage means 3 which stores the feature pattern of the a standard speech signal, a matching means 4 which collates the extracted features of the speech input signal with the stored feature pattern and specifies the standard speech signal corresponding to the input speech signal, a display pattern storage means 5 which stores display information corresponding to the standard speech signal, and a display means 6 which displays the display information; and the feature standard pattern storage means 3 stores the feature patterns 31 of human words and feature patterns 32 other than the human words and the display pattern storage means 5 stores word display information 51 generated by representing the standard speech signal with words as they are and non-word display information 52 corresponding to a non-word standard speech signal.