FontGAN: A Unified Generative Framework for Chinese Character Stylization and De-stylization

TL;DR

FontGAN is a unified generative framework that improves Chinese character stylization and de-stylization by decoupling style and content representations, ensuring better control and quality of generated characters.

Contribution

The paper introduces FontGAN, a novel model that integrates stylization and de-stylization, with modules for style consistency and content prior, enhancing control and quality in Chinese character synthesis.

Findings

Effective in character stylization and de-stylization

Improves style control and content preservation

Demonstrates superior results on experimental datasets

Abstract

Chinese character synthesis involves two related aspects, i.e., style maintenance and content consistency. Although some methods have achieved remarkable success in synthesizing a character with specified style from standard font, how to map characters to a specified style domain without losing their identifiability remains very challenging. In this paper, we propose a novel model named FontGAN, which integrates the character stylization and de-stylization into a unified framework. In our model, we decouple character images into style representation and content representation, which facilitates more precise control of these two types of variables, thereby improving the quality of the generated results. We also introduce two modules, namely, font consistency module (FCM) and content prior module (CPM). FCM exploits a category guided Kullback-Leibler loss to embedding the style representation into different Gaussian distributions. It constrains the characters of the same font in the training set globally. On the other hand, it enables our model to obtain style variables through sampling in testing phase. CPM provides content prior for the model to guide the content encoding process and alleviates the problem of stroke deficiency during de-stylization. Extensive experimental results on character stylization and de-stylization have demonstrated the effectiveness of our method.