Mutual-Learning Improves End-to-End Speech Translation

Jiawei Zhao, Wei Luo, Boxing Chen, Andrew Gilman


Abstract
A currently popular research area in end-to-end speech translation is the use of knowledge distillation from a machine translation (MT) task to improve the speech translation (ST) task. However, such scenario obviously only allows one way transfer, which is limited by the performance of the teacher model. Therefore, We hypothesis that the knowledge distillation-based approaches are sub-optimal. In this paper, we propose an alternative–a trainable mutual-learning scenario, where the MT and the ST models are collaboratively trained and are considered as peers, rather than teacher/student. This allows us to improve the performance of end-to-end ST more effectively than with a teacher-student paradigm. As a side benefit, performance of the MT model also improves. Experimental results show that in our mutual-learning scenario, models can effectively utilise the auxiliary information from peer models and achieve compelling results on Must-C dataset.
Anthology ID:
2021.emnlp-main.325
Volume:
Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
Month:
November
Year:
2021
Address:
Online and Punta Cana, Dominican Republic
Editors:
Marie-Francine Moens, Xuanjing Huang, Lucia Specia, Scott Wen-tau Yih
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3989–3994
Language:
URL:
https://aclanthology.org/2021.emnlp-main.325
DOI:
10.18653/v1/2021.emnlp-main.325
Bibkey:
Cite (ACL):
Jiawei Zhao, Wei Luo, Boxing Chen, and Andrew Gilman. 2021. Mutual-Learning Improves End-to-End Speech Translation. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 3989–3994, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics.
Cite (Informal):
Mutual-Learning Improves End-to-End Speech Translation (Zhao et al., EMNLP 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.emnlp-main.325.pdf
Video:
 https://aclanthology.org/2021.emnlp-main.325.mp4
Data
MuST-C