Blind extraction of dominant target sources

This page demonstrates sound examples for blind extraction of dominant target sources.

Experimental setups

Sound examples

Single-target cases

1 dominant target speech + 6 background interference speeches

Target position a120 b120 c120 c170

Mixture x1 x1 x1 x1

Separation only with ICA y1 y1 y1 y1

Separation by ICA and T-F masking (fewer zeros) y1 y1 y1 y1

Separation by ICA and T-F masking (moderate) y1 y1 y1 y1

Separation by ICA and T-F masking (more zeros) y1 y1 y1 y1

Multi-target case

3 dominant target speeches + 6 background interference speeches

Target position a120 b120 c170

Mixture x1

Separation only with ICA y1 y2 y3

Separation by ICA and T-F masking (moderate) y1 y2 y3

Target position	a120	b120	c120	c170
Mixture	x1	x1	x1	x1
Separation only with ICA	y1	y1	y1	y1
Separation by ICA and T-F masking (fewer zeros)	y1	y1	y1	y1
Separation by ICA and T-F masking (moderate)	y1	y1	y1	y1
Separation by ICA and T-F masking (more zeros)	y1	y1	y1	y1