Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Speech Separation on VoxCeleb2+DCASE 2 speakers+noise (test)
Loading...
7.92
SISDR
RAVSS
-3
-0.165
2.67
5.505
Feb 1, 2026
SISDR
PESQ
ESTOI
NISQA
WER (Non-Speech)
WER (Speech)
Updated 4d ago
Evaluation Results
Method
Method
Links
SISDR
PESQ
ESTOI
NISQA
WER (Non-Speech)
WER (Speech)
RAVSS
Unsupervised=false
2026.02
7.92
2.33
0.62
1.97
0.243
0.224
FLOWAVSE
Unsupervised=false
2026.02
6.8
2.07
0.61
3.51
0.295
0.269
SSNAPS
Unsupervised=true
2026.02
6.35
2.09
0.61
2.55
0.227
0.216
MIXTURE
2026.02
-2.58
1.51
0.35
1.06
0.919
0.789
Feedback
Search any
task
Search any
task