Steering Generative Models for Accessibility: EasyRead Image Generation

2026-03-13

AI Summary

This work addresses the limitations of current diffusion models in generating pictograms for accessible visual content, which often suffer from excessive detail and inconsistent styles, failing to meet the essential requirements of simplicity and uniformity. In contrast, conventional EasyRead pictograms rely on manual design, resulting in high costs and low efficiency. To overcome these challenges, the authors curate and enhance a multi-source pictogram dataset and develop a unified generation pipeline by fine-tuning Stable Diffusion with LoRA. They further introduce, for the first time, a quantifiable EasyRead scoring mechanism to evaluate output quality. Experimental results demonstrate that the proposed method significantly improves both the simplicity of generated images and their consistency across different random seeds, thereby validating the feasibility and practical potential of diffusion models for low-cost, large-scale production of accessible pictograms.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Humans and AI: AI for Accessibility
• Natural Language Processing: Generation

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

EasyRead pictograms are simple, visually clear images that represent specific concepts and support comprehension for people with intellectual disabilities, low literacy, or language barriers. The large-scale production of EasyRead content has traditionally been constrained by the cost and expertise required to manually design pictograms. In contrast, automatic generation of such images could significantly reduce production time and cost, enabling broader accessibility across digital and printed materials. However, modern diffusion-based image generation models tend to produce outputs that exhibit excessive visual detail and lack stylistic stability across random seeds, limiting their suitability for clear and consistent pictogram generation. This challenge highlights the need for methods specifically tailored to accessibility-oriented visual content. In this work, we present a unified pipeline for generating EasyRead pictograms by fine-tuning a Stable Diffusion model using LoRA adapters on a curated corpus that combines augmented samples from multiple pictogram datasets. Since EasyRead pictograms lack a unified formal definition, we introduce an EasyRead score to benchmark pictogram quality and consistency. Our results demonstrate that diffusion models can be effectively steered toward producing coherent EasyRead-style images, indicating that generative models can serve as practical tools for scalable and accessible pictogram production.

Problem

• EasyRead
• pictogram generation
• accessibility
• diffusion models
• visual consistency

Innovation

• EasyRead pictograms
• diffusion models
• LoRA fine-tuning
• accessibility-oriented generation
• visual consistency