Single channel speech enhancement using spectral gain shaping method and dereverberation for digital hearing aid

Speech enhancement algorithms are used to remove back ground noise in a speech signal and are become an integral component of modern hearing aids. The existing speech enhancement algorithms have short comings such as low speech quality, high signal delay and computational complexity. So in this paper a speech enhancement algorithm for the joint suppression of late reverberant speech and background noise which addresses the special requirements of digital hearing aids is proposed. It is based on a generalized spectral subtraction and employs a low delay spectral gain shaping method with parallel equalization filters for noise reduction. This method provides better speech intelligibility with increased SNR and with minimum Mean Square Error.