A Multi-Stage Approach to Image Consistency in Zero-Shot Character Art Generation for the D&D Domain

Gayashan Weerasundara, Nisansa de Silva

2024

Abstract

In the evolving landscape of computer graphics, the pursuit of consistency in image generation remains a compelling challenge. This paper delineates a comprehensive methodology that harnesses the capabilities of the Stable Diffusion (SD) model, an adept open-source image generation tool, to generate consistent and high-fidelity imagery. Beginning with the training of a generalized art style for the popular tabletop role-playing game Dungeons and Dragons, our research journeyed through specific character generation and the creation of exhaustive character sheets, culminating in the generation of consistent character images across various poses.

Paper Citation

in Harvard Style

Weerasundara G. and de Silva N. (2024). A Multi-Stage Approach to Image Consistency in Zero-Shot Character Art Generation for the D&D Domain. In Proceedings of the 16th International Conference on Agents and Artificial Intelligence - Volume 3: ICAART; ISBN 978-989-758-680-4, SciTePress, pages 235-242. DOI: 10.5220/0012317600003636

in Bibtex Style

@conference{icaart24,
author={Gayashan Weerasundara and Nisansa de Silva},
title={A Multi-Stage Approach to Image Consistency in Zero-Shot Character Art Generation for the D&D Domain},
booktitle={Proceedings of the 16th International Conference on Agents and Artificial Intelligence - Volume 3: ICAART},
year={2024},
pages={235-242},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012317600003636},
isbn={978-989-758-680-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 16th International Conference on Agents and Artificial Intelligence - Volume 3: ICAART
TI - A Multi-Stage Approach to Image Consistency in Zero-Shot Character Art Generation for the D&D Domain
SN - 978-989-758-680-4
AU - Weerasundara G.
AU - de Silva N.
PY - 2024
SP - 235
EP - 242
DO - 10.5220/0012317600003636
PB - SciTePress