Speech enhancement using perceptual Wiener filter combined with unvoiced speech — A new Scheme

A new speech enhancement technique is proposed that reduces musical noise and improves intelligibility of speech. The proposed system has two stages. First stage consists of Wiener filter with self adaptive averaging factor to estimate apriori SNR. To reduce residual musical noise that remains in the enhanced speech, perceptual weighting filter is employed based on simultaneous and temporal masking effects of human auditory system is used in the second stage. An unvoiced speech enhancement algorithm is also integrated in the scheme to improve the intelligibility of speech. Through rigorous objective and subjective evaluations, it is observed that the proposed speech enhancement scheme is capable of reducing noise with little speech distortion in adverse noise environments and the overall performance is superior to several other methods available in literature.