SYSTEM FOR ADAPTIVE ENHANCEMENT OF SPEECH SIGNALS

摘要

A system for enhancing the adaptability of a speech signal is provided to add the correction components of spectrum into the spectrum of an input sound signal. An input sound signal is received(102). The analysis of the frequency band is performed at the inputted signal(104). The compressed dB spectrum generated from the sub-band analysis is applied to the SNR(Signal to Noise Ratio) estimation and the sound detection(106). The dB value of the frequency band having high SNR and audio information is used for updating the mean sound spectrum type estimation(108). The mean sound spectrum type estimation is compared with the object sound spectrum type(112). The sound spectrum type correction factor is determined by subtraction of the mean sound spectrum type from the object sound spectrum type(114).