SPEECH ENHANCEMENT WITH VOICE CLARITY Russian patent published in 2012 - IPC G10L21/02 G10L11/02 

FIELD: information technology.

SUBSTANCE: in the method of enhancing speech components of an audio signal consisting of speech and noise components, the audio signal is changed from the time domain to a plurality of frequency-domain subbands, with generation of multiple subband signals; the audio signal subbands are processed, wherein said processing includes controlling the gain of the audio signal in some of the subbands, wherein the gain in a subband is controlled by additive/subtractive or multiplicative combination a) reducing gain in the subband while increasing the estimate of the level of noise components in the subband. The estimate of the noise components in the subband is determined in the absence of speech, and b) increasing gain in the subband in the presence of speech components in the audio signal subband, wherein the increase in gain is lowered in accordance with the time smoothing when switching from the presence of speech components to absence of speech components. Further, the processed audio subband audio signal is changed from the frequency domain to the time domain to provide an audio signal in which speech components are enhanced.

EFFECT: enhanced speech and voice clarity.

4 cl, 5 dwg