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 “5.3 Phoneme Recognition on TIMIT” · whole deck
5.3 Phoneme Recognition on TIMIT
- How did wav2vec 2.0 do on TIMIT?New state of the art without a language model: 7.4/8.3 phoneme error rate on dev/test, a 23%/29% relative reduction over the next best.
- Baevski et al., 2020, wav2vec 2.0 (Advances in Neural Information Processing Systems 33 (NeurIPS 2020)), §5.3
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