Speech Separation On Ikala

NSDR

Results

Performance results of various models on this benchmark

Comparison Table

Model Name	NSDR
singing-voice-separation-with-deep-u-net	11.094 (Vocal); 14.435 (Instrumental)