H∞ filtering for speech enhancement

A new approach based on H_∞ filtering is presented for speech enhancement. This approach differs from the traditional modified Wiener/Kalman filtering approach in the following two aspects: 1) no a priori knowledge of the noise statistics is required; instead the noise signals are only assumed to have finite energy; 2) the estimation criterion for the filter design is to minimize the worst possible amplification of the estimation error signal in terms of the modeling errors and additive noises. Since most additive noises in speech are not Gaussian, this approach is highly robust and is more appropriate in practical speech enhancement. The global signal-to-noise ratio (SNR), time domain speech representation and listening evaluations are used to verify the performance of the H_∞ filtering algorithm. Experimented results show that the filtering performance is better than other speech enhancement approaches in the literature under similar experimental conditions