Spectral Subtraction Speech Enhancement Technology Based on Fast Noise Estimation
Jun Luo, Zhiming He · 2009
In this paper, a spectral subtraction speech enhancement technology based on fast noise estimation is studied. Experiments show that this algorithm can significantly improve the performance compared to the basic spectral subtraction algorithm. This algorithm can estimate the noise power spectrum quickly in nonstationary noise environment and it has an easy implementation and small amount of calculation.