CustomText: Customized Textual Image Generation using Diffusion Models

TL;DR

CustomText improves textual image generation by enabling precise control over font attributes and enhancing small-text rendering accuracy using diffusion models, surpassing previous methods in quality and control.

Contribution

The paper introduces CustomText, a novel approach that leverages pre-trained diffusion models and ControlNet to improve text rendering and customization in generated images.

Findings

Superior text rendering quality on CTW-1500 dataset

Enhanced control over font color, background, and types

Significant improvement in small-text generation accuracy

Abstract

Textual image generation spans diverse fields like advertising, education, product packaging, social media, information visualization, and branding. Despite recent strides in language-guided image synthesis using diffusion models, current models excel in image generation but struggle with accurate text rendering and offer limited control over font attributes. In this paper, we aim to enhance the synthesis of high-quality images with precise text customization, thereby contributing to the advancement of image generation models. We call our proposed method CustomText. Our implementation leverages a pre-trained TextDiffuser model to enable control over font color, background, and types. Additionally, to address the challenge of accurately rendering small-sized fonts, we train the ControlNet model for a consistency decoder, significantly enhancing text-generation performance. We assess the performance of CustomText in comparison to previous methods of textual image generation on the publicly available CTW-1500 dataset and a self-curated dataset for small-text generation, showcasing superior results.