Conformer + Wav2vec 2.0 + SpecAugment-based Noisy Student Training with Libri-Light
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
- Word Error Rate (WER)· uses extra data· 2020-10-20SOTA1.4best: 0.985 (United Med ASR)
- Word Error Rate (WER)· 2020-10-20SOTA2.6best: 2.48 (SAMBA ASR)