UTTERED SOUND DETECTION APPARATUS

摘要

An apparatus (1) receives an inputted sound signal (sound receiving unit (18)) and calculates input power indicating the magnitude of sound represented by the sound signal for each frequency (input power calculation unit (11)).  The apparatus (1) estimates a correction function which is a continuous function for defining the relation between a frequency and a correction function used for approximating the input power calculated with respect to the frequency to the reference power determined with respect to the frequency (correction function estimating unit (14)).  The apparatus (1) corrects the input power for each frequency on the basis of the correction function obtained according to the relation defined by the estimated correction function (input power correction unit (12)).  On the basis of the corrected input power, the apparatus (1) determines if the sound represented by the received sound signal is an uttered sound (uttered sound detection unit (16)).