Fig. 3From: Combination of MVDR beamforming and single-channel spectral processing for enhancing noisy and reverberant speechPower of anechoic speech, reverberant speech, and additive noise at a frequency of 500 Hz for a 1-s signal extracted from the REVERB challenge corpus for a room of T 60=0.73 s and a distance of 2 m between the speech source and the microphoneBack to article page