StrokeGAN: Reducing Mode Collapse in Chinese Font Generation via Stroke Encoding

by   Jinshan Zeng, et al.

The generation of stylish Chinese fonts is an important problem involved in many applications. Most of existing generation methods are based on the deep generative models, particularly, the generative adversarial networks (GAN) based models. However, these deep generative models may suffer from the mode collapse issue, which significantly degrades the diversity and quality of generated results. In this paper, we introduce a one-bit stroke encoding to capture the key mode information of Chinese characters and then incorporate it into CycleGAN, a popular deep generative model for Chinese font generation. As a result we propose an efficient method called StrokeGAN, mainly motivated by the observation that the stroke encoding contains amount of mode information of Chinese characters. In order to reconstruct the one-bit stroke encoding of the associated generated characters, we introduce a stroke-encoding reconstruction loss imposed on the discriminator. Equipped with such one-bit stroke encoding and stroke-encoding reconstruction loss, the mode collapse issue of CycleGAN can be significantly alleviated, with an improved preservation of strokes and diversity of generated characters. The effectiveness of StrokeGAN is demonstrated by a series of generation tasks over nine datasets with different fonts. The numerical results demonstrate that StrokeGAN generally outperforms the state-of-the-art methods in terms of content and recognition accuracies, as well as certain stroke error, and also generates more realistic characters.


page 6

page 7


StrokeGAN+: Few-Shot Semi-Supervised Chinese Font Generation with Stroke Encoding

The generation of Chinese fonts has a wide range of applications. The cu...

SGCE-Font: Skeleton Guided Channel Expansion for Chinese Font Generation

The automatic generation of Chinese fonts is an important problem involv...

Understanding Deep Generative Models with Generalized Empirical Likelihoods

Understanding how well a deep generative model captures a distribution o...

Text Generation with Deep Variational GAN

Generating realistic sequences is a central task in many machine learnin...

Coconditional Autoencoding Adversarial Networks for Chinese Font Feature Learning

In this work, we propose a novel framework named Coconditional Autoencod...

Refining Deep Generative Models via Wasserstein Gradient Flows

Deep generative modeling has seen impressive advances in recent years, t...

Optimizing the Learning Order of Chinese Characters Using a Novel Topological Sort Algorithm

We present a novel algorithm for optimizing the order in which Chinese c...

Please sign up or login with your details

Forgot password? Click here to reset