Rethinking Style Transformer with Energy-based Interpretation: Adversarial Unsupervised Style Transfer using a Pretrained Model

Hojun Cho, Dohee Kim, Seungwoo Ryu, ChaeHun Park, Hyungjong Noh, Jeong-in Hwang, Minseok Choi, Edward Choi, Jaegul Choo


Abstract
Style control, content preservation, and fluency determine the quality of text style transfer models. To train on a nonparallel corpus, several existing approaches aim to deceive the style discriminator with an adversarial loss. However, adversarial training significantly degrades fluency compared to the other two metrics. In this work, we explain this phenomenon using energy-based interpretation, and leverage a pretrained language model to improve fluency. Specifically, we propose a novel approach which applies the pretrained language model to the text style transfer framework by restructuring the discriminator and the model itself, allowing the generator and the discriminator to also take advantage of the power of the pretrained model. We evaluated our model on three public benchmarks GYAFC, Amazon, and Yelp and achieved state-of-the-art performance on the overall metrics.
Anthology ID:
2022.emnlp-main.366
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5452–5467
Language:
URL:
https://aclanthology.org/2022.emnlp-main.366
DOI:
10.18653/v1/2022.emnlp-main.366
Bibkey:
Cite (ACL):
Hojun Cho, Dohee Kim, Seungwoo Ryu, ChaeHun Park, Hyungjong Noh, Jeong-in Hwang, Minseok Choi, Edward Choi, and Jaegul Choo. 2022. Rethinking Style Transformer with Energy-based Interpretation: Adversarial Unsupervised Style Transfer using a Pretrained Model. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 5452–5467, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Rethinking Style Transformer with Energy-based Interpretation: Adversarial Unsupervised Style Transfer using a Pretrained Model (Cho et al., EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.366.pdf