Table 5 Comparison of the separation approaches based on the six-channel UCA for different subtending angles
From: A multichannel learning-based approach for sound source separation in reverberant environments
Model | ΔSI-SNR (dB) | ΔPESQ | ΔSTOI | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
0°–15° | 15°–45° | 45°–90° | 90°–180° | Avg. | 0°–15° | 15°–45° | 45°–90° | 90°–180° | Avg. | Avg. | |
WPE + MPDR | 3.27 | 3.05 | 3.98 | 6.17 | 3.94 | − 1.22 | − 1.29 | − 1.31 | − 1.18 | − 1.27 | 0.09 |
WPE + TIKR | − 5.68 | − 0.92 | 4.89 | 7.61 | 2.05 | − 0.30 | − 1.03 | − 1.25 | − 1.20 | − 1.06 | 0.11 |
WPE + IVA | 3.80 | 4.10 | 4.94 | 5.98 | 4.69 | 0.58 | 0.64 | 0.77 | 0.81 | 0.71 | 0.17 |
Beam-TasNet | 7.03 | 7.55 | 7.79 | 7.48 | 7.57 | − 0.01 | 0.01 | 0.13 | 0.08 | 0.05 | 0.06 |
BF-net + U-net + DM | 14.32 | 14.27 | 13.69 | 13.81 | 13.99 | 0.51 | 0.78 | 0.95 | 0.89 | 0.83 | 0.22 |