Deep learning for Chinese font generation: a survey

Fonts, collections of characters with unique styles, directly influence the visual appearance of text. However, font design and development require extensive expertise and significant labor. In recent years, breakthroughs in deep learning have significantly advanced Chinese font generation (CFG) and its related applications. This paper provides a comprehensive review of deep learning-based Chinese font generation techniques and the latest advancements in their applications. Based on an extensive literature review, we first define the fundamental problems and key challenges of font generation. Next, we summarize publicly available datasets and evaluation metrics, analyzing their performance in practical applications. We then review the development of Chinese font generation research in the deep learning era, systematically analyzing the current state of the technology from multiple perspectives, including neural network architectures, task paradigms, and training strategies. Finally, this paper discusses the current state and challenges of research applications while outlining future research directions. This review aims to advance CFG research by offering a comprehensive perspective and inspiring future advancements.

Text 124221.pdf - Published Version Restricted to Repository staff only Download (4MB) Official URL or Download Paper: https://www.sciencedirect.com/science/article/pii/...

Actions (login required)

View Item