Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-talker Automatic Speech Recognition on Libri2Mix clean (test)
Loading...
3.1
WER
ID-30.D
2.624
5.837
9.05
12.263
Mar 28, 2026
WER
Updated 19d ago
Evaluation Results
Method
Method
Links
WER
ID-30.D
LLM usage=With LLMs
2026.03
3.1
SOP-Llama-3B
LLM usage=With LLMs, B...
2026.03
3.6
ID-20.D
LLM usage=With LLMs
2026.03
3.9
SOP-Llama-1B
LLM usage=With LLMs, B...
2026.03
4
SOT-Llama-3B
LLM usage=With LLMs, B...
2026.03
4.1
SOT-Llama-1B
LLM usage=With LLMs, B...
2026.03
4.6
GEncSep w/o decoder
LLM usage=Without LLMs...
2026.03
6.1
UME
LLM usage=Without LLMs...
2026.03
6.4
GEncSep
LLM usage=Without LLMs...
2026.03
6.6
Training from Scratch
LLM usage=Without LLMs...
2026.03
7
WavLM-CLN
LLM usage=Without LLMs...
2026.03
7.6
C-HuBERT LARGE
LLM usage=Without LLMs...
2026.03
7.8
W2V-Sidecar-ft.
LLM usage=Without LLMs...
2026.03
8.1
HCM
LLM usage=Without LLMs...
2026.03
8.2
MUSE-TSASR
LLM usage=Without LLMs...
2026.03
11
SoloSpeech
LLM usage=Without LLMs...
2026.03
15
Feedback
Search any
task
Search any
task