Abstract | ||
---|---|---|
Pseudo-labeling has recently shown promise in end-to-end automatic speech recognition (ASR). We study Iterative Pseudo-Labeling (IPL), a semi-supervised algorithm which efficiently performs multiple iterations of pseudo-labeling on unlabeled data as the acoustic model evolves. In particular, IPL fine-tunes an existing model at each iteration using both labeled data and a subset of unlabeled data. We study the main components of IPL: decoding with a language model and data augmentation. We then demonstrate the effectiveness of IPL by achieving state-of-the-art word-error rate on the Librispeech test sets in both standard and low-resource setting. We also study the effect of language models trained on different corpora to show IPL can effectively utilize additional text. Finally, we release a new large in-domain text corpus which does not overlap with the Librispeech training transcriptions to foster research in low-resource, semi-supervised ASR |
Year | DOI | Venue |
---|---|---|
2020 | 10.21437/Interspeech.2020-1800 | INTERSPEECH |
DocType | Citations | PageRank |
Conference | 6 | 0.42 |
References | Authors | |
0 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Qiantong Xu | 1 | 34 | 7.42 |
Tatiana Likhomanenko | 2 | 24 | 5.47 |
Jacob Kahn | 3 | 20 | 2.38 |
Awni Y. Hannun | 4 | 517 | 27.54 |
Synnaeve Gabriel | 5 | 21 | 5.12 |
Ronan Collobert | 6 | 4002 | 308.61 |