Model-Based Noise PSD Estimation from Speech in Non-Stationary Noise

Jesper Kjcer Nielsen, Mathew Shaji Kavalekalam, Mads Grcesboll Christensen, Jesper Boldt
2018 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
Most speech enhancement algorithms need an estimate of the noise power spectral density (PSD) to work. In this paper, we introduce a model-based framework for doing noise PSD estimation. The proposed framework allows us to include prior spectral information about the speech and noise sources, can be configured to have zero tracking delay, and does not depend on estimated speech presence probabilities. This is in contrast to other noise PSD estimators which often have a too large tracking delay
more » ... o give good results in nonstationary situations and offer no consistent way of including prior information about the speech or the noise type. The results show that the proposed method outperforms state-of-the-art noise PSD estimators in terms of tracking speed and estimation accuracy.
doi:10.1109/icassp.2018.8461683 dblp:conf/icassp/NielsenKCB18 fatcat:r2pgt6ixfff27olbzhu5nzsery