Multi-Content GAN for Few-Shot Font Style Transfer

TL;DR

This paper introduces a novel multi-content GAN model that effectively transfers complex font styles from few examples to generate consistent stylized glyphs, addressing the challenge of few-shot font style transfer.

Contribution

The paper presents an end-to-end stacked conditional GAN that captures both typographic and textual styles for few-shot font style transfer, a novel approach in this domain.

Findings

Effective style transfer from few examples demonstrated on 10,000 fonts.

The model generalizes well to highly stylized fonts like posters and infographics.

Achieves consistent multi-content image generation with minimal input data.

Abstract

In this work, we focus on the challenge of taking partial observations of highly-stylized text and generalizing the observations to generate unobserved glyphs in the ornamented typeface. To generate a set of multi-content images following a consistent style from very few examples, we propose an end-to-end stacked conditional GAN model considering content along channels and style along network layers. Our proposed network transfers the style of given glyphs to the contents of unseen ones, capturing highly stylized fonts found in the real-world such as those on movie posters or infographics. We seek to transfer both the typographic stylization (ex. serifs and ears) as well as the textual stylization (ex. color gradients and effects.) We base our experiments on our collected data set including 10,000 fonts with different styles and demonstrate effective generalization from a very small number of observed glyphs.