Figure 11
From: Monaural Voiced Speech Segregation Based on Dynamic Harmonic Function
SNR results using IBM as the ground truth. White bars show the results from unprocessed mixtures, black bars those from the Hu and Wang model, and gray bars those from proposed system.