LSTM based AE-DNN constraint for better late reverb suppression in multi-channel LP formulation
This work addresses speech enhancement for applications in noisy or reverberant environments, but it is incremental as it builds on existing MCLP methods with a DNN-based improvement.
The paper tackled the problem of late reverberation suppression in speech enhancement by proposing a deep neural network (DNN) based non-linear estimate for the desired signal power spectral density (PSD) in multi-channel linear prediction (MCLP), using an auto-encoder trained on clean speech; experiments showed that the LSTM-DNN based PSD estimate performed better than traditional methods.
Prediction of late reverberation component using multi-channel linear prediction (MCLP) in short-time Fourier transform (STFT) domain is an effective means to enhance reverberant speech. Traditionally, a speech power spectral density (PSD) weighted prediction error (WPE) minimization approach is used to estimate the prediction filters. The method is sensitive to the estimate of the desired signal PSD. In this paper, we propose a deep neural network (DNN) based non-linear estimate for the desired signal PSD. An auto encoder trained on clean speech STFT coefficients is used as the desired signal prior. We explore two different architectures based on (i) fully-connected (FC) feed-forward, and (ii) recurrent long short-term memory (LSTM) layers. Experiments using real room impulse responses show that the LSTM-DNN based PSD estimate performs better than the traditional methods for late reverb suppression.