ResNet + BiLSTMs acoustic model

Reported on 2 benchmarks across 1 task · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Audio2 results

Speech Recognitiononswb_hub_500 WER fullSWBCH
Percentage error· 2017-03-06
10.3
best: 6.8 (IBM (LSTM+Conformer encoder-decoder))
SOTA
English Conversational Telephone Speech Recognition by Humans and Machines arXiv:1703.02136
Speech RecognitiononSwitchboard + Hub500
Percentage error· 2017-03-06
5.5
best: 4.3 (IBM (LSTM+Conformer encoder-decoder))
SOTA
English Conversational Telephone Speech Recognition by Humans and Machines arXiv:1703.02136