SepIt

Reported on 4 benchmarks across 1 task · 1 paper · 4 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Speech4 results

Speech SeparationonWSJ0-2mix
SI-SDRi· 2022-05-24
22.4
best: 25.1 (TF-Locoformer (L) + DM)
SOTA
SepIt: Approaching a Single Channel Speech Separation Bound arXiv:2205.11801
Speech SeparationonWSJ0-3mix
SI-SDRi· 2022-05-24
20.1
best: 23.7 (SepTDA)
SOTA
SepIt: Approaching a Single Channel Speech Separation Bound arXiv:2205.11801
Speech SeparationonLibri5Mix
SI-SDRi· uses extra data· 2022-05-24
13.7
best: 14.2 (Separate And Diffuse)
SOTA
SepIt: Approaching a Single Channel Speech Separation Bound arXiv:2205.11801
Speech SeparationonLibri10Mix
SI-SDRi· 2022-05-24
8.2
best: 9 (Separate And Diffuse)
SOTA
SepIt: Approaching a Single Channel Speech Separation Bound arXiv:2205.11801