Fig. 5From: Channel and temporal-frequency attention UNet for monaural speech enhancementComparison of the unprocessed speech spectrogram and the enhanced speech spectrogramBack to article page