Noise reduction using simultaneous masking property and SNR variation for various noise corruptions

A speech enhancement algorithm adapted by both intra- frame masking properties of the human auditory system and inter-frame SNR variation is proposed to enhance a speech signal corrupted by colored noise. Herein, we employ a gain factor adapted by the SNR variation to reduce the spectral variation over successive frames, so the effect of musical residual noise can be mitigated. In addition, the masking property of the human ears is also employed to adapt the gain factor, enabling the imperceptive residual noise with energy below the noise masking threshold to be retained. The speech distortion is therefore reduced by preserving more noisy speech signals. Experimental results show that the proposed scheme can efficiently reduce the effect of musical residual noise by rendering residual noise perceptually white.