wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
Baevski et al., 2020, wav2vec 2.0 (Advances in Neural Information Processing Systems 33 (NeurIPS 2020))
Subscriber deck — free during early access
1 cards in “5.4 Ablations” · whole deck
5.4 Ablations
- Why does wav2vec 2.0 use continuous inputs but quantized targets?Ablations showed it was best (5.4 dev WER vs 7.97 for quantized inputs and targets, 12.18 for quantized inputs with continuous targets). Continuous inputs retain information for context; continuous targets can encode speaker and background details that make the task too easy (training accuracy rose from 62% to 78%) and the representations less general.
- Baevski et al., 2020, wav2vec 2.0 (Advances in Neural Information Processing Systems 33 (NeurIPS 2020)), §5.4
Source
- Alexei Baevski, Henry Zhou, Abdelrahman Mohamed and Michael Auli. 2020. wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations. In Advances in Neural Information Processing Systems 33 (NeurIPS 2020). https://papers.nips.cc/paper/2020/hash/92d1e1eb1cd6f9fba3227870bb6d7f07-Abstract.html