Improving English to Chinese machine translation performance via adversarial training

Huaxin Pu; Donghui Guo

doi:10.1117/12.2656585

20 October 2022 Improving English to Chinese machine translation performance via adversarial training

Huaxin Pu, Donghui Guo

Proceedings Volume 12451, 5th International Conference on Computer Information Science and Application Technology (CISAT 2022); 124513I (2022) https://doi.org/10.1117/12.2656585
Event: 5th International Conference on Computer Information Science and Application Technology (CISAT 2022), 2022, Chongqing, China

Abstract

The transformer has improved slow training, the most criticized disadvantage, of RNNs and achieves fast parallelism by self-attention mechanism. Also, the transformer can be enhanced to a great depth to fully exploit the characteristics of the deep neural network model and improve the model’s accuracy, which is widely used in machine translation. Deep models, however, have the inevitable disadvantage of being prone to overfitting. We consider introducing the adversarial training method in the translation model to improve the robustness of the model through regularization to alleviate the overfitting problem. We conducted analytical experiments on the English-Chinese translation dataset which shows that the adversarial training approach can improve translation quality.

Citation Download Citation

Huaxin Pu and Donghui Guo "Improving English to Chinese machine translation performance via adversarial training", Proc. SPIE 12451, 5th International Conference on Computer Information Science and Application Technology (CISAT 2022), 124513I (20 October 2022); https://doi.org/10.1117/12.2656585

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available