Home // CONTENT 2024, The Sixteenth International Conference on Creative Content Technologies // View article
Authors:
Irina Kogai
Keywords: DALL-E 3; children’s fairy tales; automatic illustration generation; AI-generated illustrations; visual storytelling.
Abstract:
This paper explores the capabilities of DALL-E 3, an advanced artificial intelligence model, to automatically generate illustrations for children’s fairy tales. The primary objective is to assess the relevance of the images produced by this Artificial Intelligence (AI), examining how well these illustrations can complement and enhance the storytelling experience for young audiences. To achieve this, we used Toloka.ai, a crowdsourcing platform where a diverse pool of evaluators assessed the relevance of AI-generated illustrations to specific excerpts of fairy tales. This methodology allowed for a broad-based, democratized approach to understanding how illustrations created by AI are perceived in terms of accuracy, engagement, and emotional reso- nance with the source material. Our findings reveal that DALL- E 3 demonstrates a significant potential in creating visually appealing, contextually appropriate, and emotionally resonant illustrations that can rival traditional artistic methods, with a relevance rate of 62%. However, the conducted research shows that images generated by DALL-E 3 are often irrelevant and inconsistent, which makes it hard to use a system like this in a fully automated fashion without human supervision. This research contributes to the emerging field of AI in creative industries, offering insights for developers, educators, and storytellers seeking innovative approaches to enrich children’s literary experiences.
Pages: 14 to 18
Copyright: Copyright (c) IARIA, 2024
Publication date: April 14, 2024
Published in: conference
ISSN: 2308-4162
ISBN: 978-1-68558-159-6
Location: Venice, Italy
Dates: from April 14, 2024 to April 18, 2024