EURASIP Journal on Audio, Speech, and Music Processing

Table 6 Comparison of the separation approaches based on the six-channel UCA for different reverberation time

From: A multichannel learning-based approach for sound source separation in reverberant environments

Model	ΔSI-SNR (dB)				ΔPESQ				ΔSTOI
Model	0.16 s	0.36 s	0.61 s	0.9 s	0.16 s	0.36 s	0.61 s	0.9 s	Avg.
WPE + MPDR	0.00	4.16	5.69	6.25	− 1.42	− 1.33	− 1.21	− 1.10	0.09
WPE + TIKR	− 0.41	2.44	3.26	3.10	− 1.28	− 1.14	− 0.96	− 0.83	0.11
WPE + IVA	3.88	4.86	5.18	4.88	0.81	0.80	0.70	0.51	0.17
Beam-TasNet	7.37	7.68	7.79	7.42	0.25	0.10	− 0.05	− 0.10	0.06
BF-net + U-net + DM	12.33	13.78	15.02	14.98	1.10	0.90	0.70	0.59	0.22

The best scores for the respective subset are boldfaced

Back to article page