wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
This work addresses the challenge of speech recognition with limited labeled data, offering a simpler and more effective approach that could reduce reliance on costly annotations.
The paper tackles the problem of speech recognition by proposing a self-supervised learning framework that learns from speech audio alone and fine-tunes on transcribed data, achieving word error rates of 1.8/3.3 on Librispeech clean/other test sets and outperforming previous methods with 100 times less labeled data.
We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/other test sets. When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 4.8/8.2 WER. This demonstrates the feasibility of speech recognition with limited amounts of labeled data.