Method for suppressing noise in an audio signal by an algorithm with variable spectral gain with dynamically adaptive strength

摘要

The method involves generating (10) a set of successive time frames (y-k) of a digitalized noisy audio signal (y-n), and applying a Fourier transform (12) to the set of successive time frames so as to produce a signal spectrum (Y-ki) for each signal time frame. A variable spectral gain (16) is calculated for each frequency band of each current time frame. A modulatable value is calculated by applying increment/decrement to a parameterized nominal value (Gmin) of a minimal gain for the current time frame.