A new technique for street noise reduction in signal processing applications

This paper proposes a two stage hybrid speech enhancement system with nonuniform subbands. Frequency bins after Fourier transform are nonuniformly grouped to reduce the computations in calculating the spectral gain. First stage includes a soft decision gain modification and applied to the Ephraim-Malah gain function based on minimum mean square error estimation (MMSE) and a psychoacoustic masking threshold is used in the second stage noise reduction. The performance of this algorithm is compared to modified spectral subtraction and spectral weighting algorithms in terms of listening tests and spectrograms. The proposed algorithm is suitable for noise reduction in non-stationary noise environments.