Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Low Resource Style Transfer via Domain Adaptive Meta Learning

About

Text style transfer (TST) without parallel data has achieved some practical success. However, most of the existing unsupervised text style transfer methods suffer from (i) requiring massive amounts of non-parallel data to guide transferring different text styles. (ii) colossal performance degradation when fine-tuning the model in new domains. In this work, we propose DAML-ATM (Domain Adaptive Meta-Learning with Adversarial Transfer Model), which consists of two parts: DAML and ATM. DAML is a domain adaptive meta-learning approach to learn general knowledge in multiple heterogeneous source domains, capable of adapting to new unseen domains with a small amount of data. Moreover, we propose a new unsupervised TST approach Adversarial Transfer Model (ATM), composed of a sequence-to-sequence pre-trained language model and uses adversarial style training for better content preservation and style transfer. Results on multi-domain datasets demonstrate that our approach generalizes well on unseen low-resource domains, achieving state-of-the-art results against ten strong baselines.

Xiangyang Li, Xiang Long, Yu Xia, Sujian Li• 2022

Related benchmarks

TaskDatasetResultRank
Semantics preservationYelp
CTC Score46.2
7
Style TransferIMDB
Accuracy24.13
7
Style TransferAMAZON
Accuracy32.9
7
Style TransferYelp
Accuracy14.2
7
Style TransferYahoo
Accuracy35.33
7
Semantics preservationIMDB
CTC Score45.6
7
Semantics preservationAMAZON
CTC Score0.455
7
Semantics preservationYahoo
CTC Score43.7
7
Showing 8 of 8 rows

Other info

Follow for update