Fig. 8From: Deep encoder/decoder dual-path neural network for speech separation in noisy reverberation environmentsResults of the MUSHRA listening test with 95% confidence intervalsBack to article page