AI-Driven Comic Generation - Exploring the Creative Potential of Generative AI in Digital Storytelling - Master of Science Thesis

Aug 24, 2026

Upload by Sarah

Creative sectors have adopted AI for modern ways to generate digital content. This thesis presents an AI-based framework to automate comic creation, combining GANs with Stable Diffusion for integrated text-to-visual generation. It aligns narrative outlining from text with automated image synthesis to produce adaptable comic panels with appropriate visual structure. Speech-bubble placement is optimized via a dedicated algorithm to keep readability and aesthetics. Iterative training and tuning are evaluated using coherence observations and quantitative metrics such as FID for images and BLEU for text, alongside user feedback.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages111

Reading Time4 h 40 min

AccessFree to download

Introduction

Background and Motivation

Problem Statement

Overview of methodology

Layout of the thesis

Literature Review

AI in creative Industries

AI for Image Generation in Comics

AI for Text Generation in Comics

Speech Bubble Placement and Text-Image Alignment

Existing Approaches to AI-Generated Comics

Hochschule Rhein-Waal Rhine-Waal University of Applied Sciences Faculty of Communication and Environment Professor Dr.-Ing. Ido Iurgel Professor Dr. Frank Zimmer AI-Driven Comic Generation: Exploring the Creative Potential of Generative AI in Digital Storytelling A Thesis Submitted in Partial Fulfillment of the Requirements of the Degree of Master of Science in Information Engineering and Computer Science by Umair Afzal Matriculation number: 28004 Submission date: 07-07-2025 Acknowledgement For maximum appreciation, I would like to point out that my supervisors, Prof. Dr.-Ing. Ido Iurgel and Prof. Dr. Frank Zimmer demonstrated exceptional dedication and mentorship that significantly shaped this work. This work would not have come about without their encouragement and their refreshing suggestions. A cohort of friends and colleagues, whom I treasure, I thank especially. They not only constantly motivated me on this journey but also engaged with me in enriching discussions. Such great aid drummed up morale to face the compounded hurdles that this journey had to present. To my family, I owe the greatest enshrinement. They have supported me, bestowing remarkable patience during this entire quest to achieve something. As they cheered me all along, it spurred me to achieve this. Abstract Creative sectors have accepted AI into their industry for establishing modern methods of generating digital content. This study introduces a novel AI-based framework based on AI technology, using GANs along with Stable Diffusion models to automate comic development. This study looks at juxtaposing narrative outlining through text with automated visual generation toward an integrated system that produces adaptable comic panels with appropriate visual structure. The research methodology that the project has followed can be built on three cornerstones: advanced GAN schemes for text generation and pre-processing, followed by image synthesis through Stable Diffusion. A specially developed algorithm for speech bubbles determined the optimal placement of that text, so it would function well and maintain a semblance of beauty. By iteratively refining and tuning the model, this system was evaluated. Initial observations regarding visual coherence and narrative alignment were hopeful, but further tests using quantitative metrics-for instance, FID for images and BLEU for text, as well as broader user feedback-would be needed to validate the efficacy of the model entirely. The impediments were, however, GAN mode collapse, irregular speech-bubble layout, and inconsistent artistic styles. Further research would uncover the potential role of AI systems in easing the comic generation process for creators, educators, and designers of digital content to enhance accessibility and efficiency. This method shows promising applicability in various domains like automated narratives, customizable comics, and educational material. Further along the way, the scientists plan to improve panel storytelling, create an intuitive interface, and expand the dataset to include more artistic styles. Such enhancements could maximize the gain from automated comic conception. Keywords: AI-driven comics, Generative Adversarial Networks, Stable Diffusion, automated storytelling, speech bubble optimization, image synthesis, digital content creation. List of Abbreviations AI Artificial Intelligence GAN Generative Adversarial Network CNN Convolutional Neural Network LDM Latent Diffusion Model GPT Generative Pre-trained Transformer MLP Multilayer Perceptron YOLO You Only Look Once BERT Bidirectional Encoder Representations from Transformers FID Fréchet Inception Distance BLEU Bilingual Evaluation Understudy NLP Natural Language Processing OCR Optical Character Recognition EDA Exploratory Data Analysis RNN Recurrent Neural Network T5 Text-To-Text Transfer Transformer ViT Vision Transformer GNN Graph Neural Net

FAQs

What core system does the thesis propose for generating comics with AI?

The thesis proposes an integrated framework that combines GAN-based text generation and pre-processing with Stable Diffusion for image synthesis. It connects text narrative outlining to automated visual generation to create comic panels with coherent structure.

How does the system ensure proper placement of speech bubbles and readable text?

A specially developed algorithm determines the optimal placement of speech bubbles so the text works well and maintains an aesthetic appearance. This supports alignment between the text elements and the generated visuals.

What metrics and evaluation approaches are used to validate the model?

Evaluation includes qualitative observations of visual coherence and narrative alignment, and quantitative metrics such as FID for image quality and BLEU for text assessment. Broader user feedback is also suggested for fuller validation.

Aug 24, 2026

Upload by Sarah