Espresso

Reported on 5 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.

Audio5 results

Speech RecognitiononWSJ eval92
Word Error Rate (WER)· 2019-09-18
3.4
best: 1.3 (Speechstew 100M)
SOTA
Espresso: A Fast End-to-end Neural Speech Recognition Toolkit arXiv:1909.08723
Speech RecognitiononHub5'00 SwitchBoard
Eval2000· 2019-09-18
9.2
SOTA
Espresso: A Fast End-to-end Neural Speech Recognition Toolkit arXiv:1909.08723
Speech RecognitiononHub5'00 CallHome
Word Error Rate (WER)· 2019-09-18
19.1
SOTA
Espresso: A Fast End-to-end Neural Speech Recognition Toolkit arXiv:1909.08723
Speech RecognitiononLibriSpeech test-clean
Word Error Rate (WER)· 2019-09-18
2.8
best: 0.985 (United Med ASR)
Espresso: A Fast End-to-end Neural Speech Recognition Toolkit arXiv:1909.08723
Speech RecognitiononLibriSpeech test-other
Word Error Rate (WER)· 2019-09-18
8.7
best: 2.48 (SAMBA ASR)
Espresso: A Fast End-to-end Neural Speech Recognition Toolkit arXiv:1909.08723