ST^2: Small-data Text Style Transfer via Multi-task Meta-Learning

by   Xiwen Chen, et al.

Text style transfer aims to paraphrase a sentence in one style into another style while preserving content. Due to lack of parallel training data, state-of-art methods are unsupervised and rely on large datasets that share content. Furthermore, existing methods have been applied on very limited categories of styles such as positive/negative and formal/informal. In this work, we develop a meta-learning framework to transfer between any kind of text styles, including personal writing styles that are more fine-grained, share less content and have much smaller training data. While state-of-art models fail in the few-shot style transfer task, our framework effectively utilizes information from other styles to improve both language fluency and style transfer accuracy.


Multi-Pair Text Style Transfer on Unbalanced Data

Text-style transfer aims to convert text given in one domain into anothe...

Low Resource Style Transfer via Domain Adaptive Meta Learning

Text style transfer (TST) without parallel data has achieved some practi...

Multi-dimensional Style Transfer for Partially Annotated Data using Language Models as Discriminators

Style transfer has been widely explored in natural language generation w...

In-Style: Bridging Text and Uncurated Videos with Style Transfer for Text-Video Retrieval

Large-scale noisy web image-text datasets have been proven to be efficie...

From Theories on Styles to their Transfer in Text: Bridging the Gap with a Hierarchical Survey

Humans are naturally endowed with the ability to write in a particular s...

Disentangling Style and Content in Anime Illustrations

Existing methods for AI-generated artworks still struggle with generatin...

Please sign up or login with your details

Forgot password? Click here to reset