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

arXiv: https://arxiv.org/abs/2012.08687 · HTML: https://mlsanity.com/paper/2012.08687/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Dec 2020.

## Built on by (most cited first, 18 of 18)

- [Diff-Font: Diffusion Model for Robust One-Shot Font Generation](https://mlsanity.com/paper/2212.05895/) (2022)
- [GlyphDraw: Seamlessly Rendering Text with Intricate Spatial Structures in Text-to-Image Generation](https://mlsanity.com/paper/2303.17870/) (2023)
- [AGTGAN: Unpaired Image Translation for Photographic Ancient Character Generation](https://mlsanity.com/paper/2303.07012/) (2023)
- [Few-shot Font Generation with Weakly Supervised Localized Representations](https://mlsanity.com/paper/2112.11895/) (2021)
- [SE-GAN: Skeleton Enhanced GAN-based Model for Brush Handwriting Font Generation](https://mlsanity.com/paper/2204.10484/) (2022)
- [FontTransformer: Few-shot High-resolution Chinese Glyph Image Synthesis via Stacked Transformers](https://mlsanity.com/paper/2210.06301/) (2022)
- [HFH-Font: Few-shot Chinese Font Synthesis with Higher Quality, Faster Speed, and Higher Resolution](https://mlsanity.com/paper/2410.06488/) (2024)
- [Stroke-Based Autoencoders: Self-Supervised Learners for Efficient Zero-Shot Chinese Character Recognition](https://mlsanity.com/paper/2207.08191/) (2022)
- [Instance Segmentation for Chinese Character Stroke Extraction, Datasets and Benchmarks](https://mlsanity.com/paper/2210.13826/) (2022)
- [Learning Stage-wise GANs for Whistle Extraction in Time-Frequency Spectrograms](https://mlsanity.com/paper/2304.02714/) (2023)
- [Count, Decode and Fetch: A New Approach to Handwritten Chinese Character Error Correction](https://mlsanity.com/paper/2307.16253/) (2023)
- [Shapes2Toon: Generating Cartoon Characters from Simple Geometric Shapes](https://mlsanity.com/paper/2211.02141/) (2022)
- [StrokeGAN+: Few-Shot Semi-Supervised Chinese Font Generation with Stroke Encoding](https://mlsanity.com/paper/2211.06198/) (2022)
- [SGCE-Font: Skeleton Guided Channel Expansion for Chinese Font Generation](https://mlsanity.com/paper/2211.14475/) (2022)
- [GAS-NeXt: Few-Shot Cross-Lingual Font Generator](https://mlsanity.com/paper/2212.02886/) (2022)
- [Computer-Aided Multi-Stroke Character Simplification by Stroke Removal](https://mlsanity.com/paper/2506.23106/) (2025)
- [NURBS Splatting: A Unified Differentiable Rendering Framework for Vector Graphics](https://mlsanity.com/paper/2606.31764/) (2026)
- [VecFontLLM: Anchor-Guided Direct Synthesis of Chinese Vector Fonts](https://mlsanity.com/paper/2607.17251/) (2026)

## Abstract (arXiv)

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.