ContextNet: Improving Convolutional Neural Networks for Automatic Speech Recognition with Global Context

Open paper
Benchmark Model Rank Results
speech-recognition-on-librispeech-test-cleanContextNet(L)#68Word Error Rate (WER): 1.9
speech-recognition-on-librispeech-test-cleanContextNet(M)#73Word Error Rate (WER): 2
speech-recognition-on-librispeech-test-cleanContextNet(S)#97Word Error Rate (WER): 2.3
speech-recognition-on-librispeech-test-otherContextNet(L)#68Word Error Rate (WER): 4.1
speech-recognition-on-librispeech-test-otherContextNet(M)#82Word Error Rate (WER): 4.5
speech-recognition-on-librispeech-test-otherContextNet(S)#102Word Error Rate (WER): 5.5

Rank counts only results with a code link.