UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data

In this paper, we propose a unified pre-training approach called UniSpeech to learn speech representations with both labeled and unlabeled data, in which supervised phonetic CTC learning and phonetically-aware contrastive self-supervised learning are conducted in a multitask learning manner. The resultant representations can capture information more correlated with phonetic structures and improve the generalization across languages and domains. We evaluate the effectiveness of UniSpeech for cross-lingual representation learning on the public CommonVoice corpus. The results show that UniSpeech outperforms self-supervised pre-training and supervised transfer learning for speech recognition by up to 13.4% and 26.9% relative phone error rate respectively (averaged over all testing languages). The transferability of UniSpeech is also verified on a domain-shift speech recognition task, demonstrating a relative word error rate reduction of 6% against the previous approach(1).