Abstract | ||
---|---|---|
We describe a method to jointly pre-train speech and text in an encoder-decoder modeling framework for speech translation and recognition. The proposed method incorporates four self-supervised and supervised subtasks for cross modality learning. A self-supervised speech subtask leverages un-labelled speech data, and a (self-)supervised text to text subtask makes use of abundant text training data. Two auxiliary supervised speech tasks are included to unify speech and text modeling space. Our contribution lies in integrating linguistic information from the text corpus into the speech pre-training. Detailed analysis reveals learning interference among subtasks. Two pre-training configurations for speech translation and recognition, respectively, are presented to alleviate subtask interference. Our experiments show the proposed method can effectively fuse speech and text information into one model. It achieves between 1.7 and 2.3 BLEU improvement above the state of the art on the MUST-C speech translation dataset and comparable WERs to wav2vec 2.0 on the LIBRISPEECH speech recognition task. (1) |
Year | DOI | Venue |
---|---|---|
2022 | 10.18653/v1/2022.acl-long.105 | PROCEEDINGS OF THE 60TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2022), VOL 1: (LONG PAPERS) |
DocType | Volume | Citations |
Conference | Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) | 0 |
PageRank | References | Authors |
0.34 | 0 | 11 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yun Tang | 1 | 0 | 0.68 |
Hongyu Gong | 2 | 0 | 1.01 |
Ning Dong | 3 | 0 | 0.34 |
Changhan Wang | 4 | 0 | 2.37 |
Wei-Ning Hsu | 5 | 115 | 13.93 |
Jiatao Gu | 6 | 274 | 22.59 |
Alexei Baevski | 7 | 85 | 9.52 |
Xian Li | 8 | 136 | 16.76 |
Abdel-rahman Mohamed | 9 | 3772 | 266.13 |
Michael Auli | 10 | 1061 | 53.54 |
Juan Pino | 11 | 21 | 12.63 |