Multi-Microphone Complex Spectral Mapping for Speech Dereverberation
This addresses speech clarity issues in noisy environments for applications like hearing aids or communication systems, but appears incremental as it builds on existing spectral mapping and beamforming techniques.
The study tackled speech dereverberation using a multi-microphone array by training a DNN to predict direct sound components from reverberant inputs, integrating it with beamforming and post-filtering, and demonstrated effectiveness in experiments.
This study proposes a multi-microphone complex spectral mapping approach for speech dereverberation on a fixed array geometry. In the proposed approach, a deep neural network (DNN) is trained to predict the real and imaginary (RI) components of direct sound from the stacked reverberant (and noisy) RI components of multiple microphones. We also investigate the integration of multi-microphone complex spectral mapping with beamforming and post-filtering. Experimental results on multi-channel speech dereverberation demonstrate the effectiveness of the proposed approach.