Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-talker Speech Recognition on LibriSpeechMix LSM-3mix (zero-shot dev)
Loading...
19.7
WER
H-SAGE
19.536
20.643
21.75
22.857
Jul 2, 2026
WER
Updated 23d ago
Evaluation Results
Method
Method
Links
WER
H-SAGE
Parm.(M)=35.75
2026.07
19.7
GLAD-SOT
Parm.(M)=35.18
2026.07
19.9
SOT-SACTC
Parm.(M)=35.77
2026.07
20.6
SOT + Local MoLE
Parm.(M)=35.09
2026.07
21.1
SOT
Parm.(M)=36.07
2026.07
23.8
Feedback
Search any
task
Search any
task