Self-supervised learning of speech representations with wav2vec

ICML 2020