Today, speech technology is only available for a small fraction of the thousands of languages spoken around the world because traditional systems need to be trained on large amounts of annotated speech audio with transcriptions. Obtaining that kind of data for every human language and dialect is almost impossible.
Wav2vec works around this limitation by requiring little to no transcribed data. The model uses self-supervision to push the boundaries by learning from unlabeled training data. This enables speech recognition systems for many more languages and dialects, such as Kyrgyz and Swahili, which don’t have a lot of transcribed speech audio. Self-supervision is the key to leveraging unannotated data and building better systems.
D. Demner-fushman, and J. Lin. Proceedings of the 21st International Conference on Computational Linguistics and the 44th annual meeting of the ACL, page 841--848. (2006)
R. Angelova, and G. Weikum. SIGIR '06: Proceedings of the 29th annual international ACM SIGIR conference on Research and development in information retrieval, page 485--492. New York, NY, USA, ACM, (2006)
F. Beil, M. Ester, and X. Xu. Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining, page 436--442. ACM Press, (2002)
F. Beil, M. Ester, and X. Xu. Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining, page 436--442. ACM Press, (2002)
I. Dhillon, S. Mallela, and D. Modha. KDD '03: Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining, page 89--98. New York, NY, USA, ACM, (2003)