Gaussian Mixture Model for Singing Voice Separation from Stereophonic Music

Created: 2022-05-17T00:06:02-05:00

Return to the Index

This card pertains to a resource available on the internet.

Hard responsibility: select a single curve to identify a sample.

Soft responsibility: select how much each curve in the mixture model represents the sample.

Soft models where stereotypes can be merged outperform hard models which cannot do that.