Jasper DR 10x5 (+ Time/Freq Masks)

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 RecognitiononLibriSpeech test-clean
Word Error Rate (WER)· 2019-04-05
2.84
best: 0.985 (United Med ASR)
SOTA
Jasper: An End-to-End Convolutional Neural Acoustic Model arXiv:1904.03288
Speech RecognitiononLibriSpeech test-other
Word Error Rate (WER)· 2019-04-05
7.84
best: 2.48 (SAMBA ASR)
SOTA
Jasper: An End-to-End Convolutional Neural Acoustic Model arXiv:1904.03288