Speech Separation on WHAMR! 2-speaker (test)
17.3SI-SNRiTF-GridNet (L)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TF-GridNet (L)# Params (M)=14.4, GMAC/s (G/s)=231.1, Model scale=L2025.07 | 17.3 | 15.8 | |
| SepReformer (L) + DM# Params (M)=55.3, GMAC/s (G/s)=155.5, Model scale=L, dynamic mixing (+DM)=true2025.07 | 17.2 | 16 | |
| MossFormer2 + DM# Params (M)=55.7, GMAC/s (G/s)=84.2, dynamic mixing (+DM)=true2025.07 | 17 | — | |
| MossFormer (L) + DM# Params (M)=42.1, GMAC/s (G/s)=70.4, Model scale=L, dynamic mixing (+DM)=true2025.07 | 16.3 | — | |
| PRESS-12 @ 12 (M) + FT# Params (M)=22.4, GMAC/s (G/s)=79.7, Model scale=M, Exit point=12, finetuned (+FT)=true2025.07 | 15.67 | 14.43 | |
| PRESS-12 @ 8 (M) + FT# Params (M)=15.6, GMAC/s (G/s)=54.4, Model scale=M, Exit point=8, finetuned (+FT)=true2025.07 | 15.61 | 14.38 | |
| PRESS-4 @ 4 (S) + FT# Params (M)=3.4, GMAC/s (G/s)=11.3, Model scale=S, Exit point=4, finetuned (+FT)=true2025.07 | 15.13 | 13.92 | |
| PRESS-12 @ 4 (M) + FT# Params (M)=8.7, GMAC/s (G/s)=29.1, Model scale=M, Exit point=4, finetuned (+FT)=true2025.07 | 14.91 | 13.73 | |
| PRESS-12 @ 12 (M)# Params (M)=22.4, GMAC/s (G/s)=79.7, Model scale=M, Exit point=122025.07 | 14.69 | 13.47 | |
| PRESS-12 @ 8 (M)# Params (M)=15.6, GMAC/s (G/s)=54.4, Model scale=M, Exit point=82025.07 | 14.67 | 13.45 | |
| PRESS-4 @ 4 (S)# Params (M)=3.4, GMAC/s (G/s)=11.3, Model scale=S, Exit point=42025.07 | 14.54 | 13.37 | |
| PRESS-12 @ 4 (M)# Params (M)=8.7, GMAC/s (G/s)=29.1, Model scale=M, Exit point=42025.07 | 14.24 | 13.09 | |
| SepFormer# Params (M)=26.0, GMAC/s (G/s)=86.92025.07 | 14 | — | |
| SepFormer + DM# Params (M)=26.0, GMAC/s (G/s)=86.9, dynamic mixing (+DM)=true2025.07 | 14 | 13 | |
| DualPathRNN# Params (M)=2.6, GMAC/s (G/s)=42.52025.07 | 10.3 | — | |
| Conv-TasNet# Params (M)=5.1, GMAC/s (G/s)=10.52025.07 | 8.3 | — |