Skip to main content

Table 1 SNR gain (in decibels) of the proposed CASA system and related systems.

From: Musical Sound Separation Based on Binary Time-Frequency Masking

Separation methods SNR gain (dB)
Proposed system 12.3
Hu and Wang (2004) 9.1
Virtanen (2006) 11.0
Parsons (1976) 10.6
Ideal Binary Mask 15.3
2-Pitch labeling 11.3
2-Pitch labeling (ideal segmentation) 13.1
Spectral smoothness 9.8