Glyph-ByT5: A Customized Text Encoder for Accurate Visual Text Rendering

TL;DR

Glyph-ByT5 is a specialized text encoder designed to improve visual text rendering in text-to-image models, achieving near-perfect accuracy and enabling complex paragraph rendering through fine-tuning and integration with SDXL.

Contribution

The paper introduces Glyph-ByT5, a character-aware, glyph-aligned text encoder, and demonstrates its integration with SDXL to significantly enhance text rendering accuracy in design and real-world images.

Findings

Text rendering accuracy improved from <20% to nearly 90%.

Glyph-SDXL can render multi-line paragraphs with high spelling accuracy.

Fine-tuning with high-quality images boosts scene text rendering in open images.

Abstract

Visual text rendering poses a fundamental challenge for contemporary text-to-image generation models, with the core problem lying in text encoder deficiencies. To achieve accurate text rendering, we identify two crucial requirements for text encoders: character awareness and alignment with glyphs. Our solution involves crafting a series of customized text encoder, Glyph-ByT5, by fine-tuning the character-aware ByT5 encoder using a meticulously curated paired glyph-text dataset. We present an effective method for integrating Glyph-ByT5 with SDXL, resulting in the creation of the Glyph-SDXL model for design image generation. This significantly enhances text rendering accuracy, improving it from less than $20\%$ to nearly $90\%$ on our design image benchmark. Noteworthy is Glyph-SDXL's newfound ability for text paragraph rendering, achieving high spelling accuracy for tens to hundreds of characters with automated multi-line layouts. Finally, through fine-tuning Glyph-SDXL with a small set of high-quality, photorealistic images featuring visual text, we showcase a substantial improvement in scene text rendering capabilities in open-domain real images. These compelling outcomes aim to encourage further exploration in designing customized text encoders for diverse and challenging tasks.