Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Unsupervised Automatic Speech Recognition on LibriSpeech 100 hours (test-other)
Loading...
0.085
PER
REBORN
0.07912
0.11881
0.1585
0.19819
Feb 6, 2024
PER
WER
Updated 4d ago
Evaluation Results
Method
Method
Links
PER
WER
REBORN
Self-training=HMM ST
2024.02
0.085
0.137
Train from oracle
Boundary=Oracle
2024.02
0.1
0.168
REBORN
2024.02
0.125
0.187
wav2vec-U 2.0
Reproduction=true, Sel...
2024.02
0.131
0.196
wav2vec-U 2.0
Reproduction=true
2024.02
0.163
0.222
EURO
Backbone=HuBERT
2024.02
0.211
0.298
wav2vec-U
Reproduction=true
2024.02
0.232
0.258
Feedback
Search any
task
Search any
task