Gated DualPathRNN

Reported on 4 benchmarks across 1 task · 1 paper · 3 SOTA

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

Speech4 results

Speech SeparationonWSJ0-5mix
SI-SDRi· 2020-02-29
10.56
best: 21 (SepTDA)
SOTA
Voice Separation with an Unknown Number of Multiple Speakers arXiv:2003.01531
Speech SeparationonWSJ0-3mix
SI-SDRi· 2020-02-29
16.85
best: 23.7 (SepTDA)
SOTA
Voice Separation with an Unknown Number of Multiple Speakers arXiv:2003.01531
Speech SeparationonWSJ0-4mix
SI-SDRi· 2020-02-29
12.88
best: 22 (SepTDA)
SOTA
Voice Separation with an Unknown Number of Multiple Speakers arXiv:2003.01531
Speech SeparationonWSJ0-2mix
SI-SDRi· 2020-02-29
20.12
best: 25.1 (TF-Locoformer (L) + DM)
Voice Separation with an Unknown Number of Multiple Speakers arXiv:2003.01531