Creación de ambientes para niveles de videojuegos utilizando Stable Diffusion

Pedro Meira-Rodríguez

doi:10.31637/epsir-2025-1354

Authors

Pedro Meira-Rodríguez Universidade da Coruña https://orcid.org/0009-0000-8355-5324

DOI:

https://doi.org/10.31637/epsir-2025-1354

Keywords:

setdressing, video game environments, level design, Stable Diffusion, ControlNet, artificial intelligence, diffusion models, digital art

Abstract

Introduction: This study investigates the optimization of time in video game development by evaluating the capability of the Stable Diffusion model to assist in set dressing. Methodology: A comparative study was conducted using three images of the same three-dimensional environment, performing iterative requests in Stable Diffusion XL with ControlNet and ConfyUI extensions, and a LoRA to unify the style. Results: The evaluation considered formal creativity, adherence to spatial design, and detail definition, resulting in average scores per indication. Discussions: The study demonstrated that the use of specific variables and weights produces images with predictable qualities, confirming a direct correlation between input parameters and outcomes. Conclusions: The importance of precisely controlling these variables was highlighted, and the utility of AI tools was emphasized. However, it is recommended to define a clear usage pathway for their effective integration into video game production.

Downloads

Download data is not yet available.

Author Biography

Pedro Meira-Rodríguez, Universidade da Coruña

Graduated in architecture studies in 2021 from the Universidade da Coruña. Master in Video Game Design and Development in 2022 from the Complutense University of Madrid. PhD student in the Official Doctorate Program in New Perspectives in Documentation, Communication and Humanities at the Universidade da Coruña since 2023. Teaching and research staff at the Universidade da Coruña in the Department of Civil Engineering (September 2023-present).

References

Abdelsalam Soliman, M. (abril de 2024). LinkedIn post April 2024. LinkedIn. https://www.linkedin.com/posts/mai-abdelsalam_comfyui-grasshopper-ai-activity-7180160257367629824-QtCZ/

Ahmad, N. B., Barakji, S. A. R., Shahada, T. M. A., y Anabtawi, Z. A. (2017). How to launch a successful video game: A framework. Entertainment Computing, 23, 1-11. https://doi.org/10.1016/j.entcom.2017.08.001 DOI: https://doi.org/10.1016/j.entcom.2017.08.001

Ameneh, P. y Microsoft, G. (2023). LoRA-Enhanced Distillation on Guided Diffusion Models. https://arxiv.org/abs/2312.06899v1

Andrew. (16 de noviembre de 2023). Conoce estos parámetros importantes para imágenes impresionantes de IA - Arte de difusión estable. https://stable-diffusion-art.com/know-these-important-parameters-for-stunning-ai-images/#CFG_Scale

Andrew. (19 de marzo de 2024). Stable Diffusion Models: a beginner’s guide. Stable Diffusion Art. https://stable-diffusion-art.com/models/

Andrew. (2023a). How to use Stable Diffusion to create AI-generated images. STABLE DIFFUSION ART. https://stable-diffusion-art.com/beginners-guide/

Bulut, E. (2023). The Fantasy of Do What You Love and Ludic Authoritarianism in the Videogame Industry. Television & New Media, 24(8), 851-869. DOI: https://doi.org/10.1177/15274764231156377

Chang, M., Druga, S., Fiannaca, A. J., Vergani, P., Kulkarni, C., Cai, C. J. y Terry, M. (2023). The Prompt Artists. Proceedings of the 15th Conference on Creativity and Cognition, 75–87. https://doi.org/10.1145/3591196.3593515 DOI: https://doi.org/10.1145/3591196.3593515

del Campo, M. (2021). Architecture, Language and Ai. Proceedings of the 26th International Conference of the Association for Computer-Aided Architectural Design Research in Asia, Volumen 1, 1, pp. 211–220. DOI: https://doi.org/10.52842/conf.caadria.2021.1.211

del Campo, M. y Leach, N. (2022). Unleashing New Creativities. Architectural Design, 92(3), 122-135. https://doi.org/10.1002/ad.2823 DOI: https://doi.org/10.1002/ad.2823

del Campo, M., Carlson, A. y Manninger, S. (2020). How machines learn to plan: A critical interrogation of machine vision techniques in architecture. Proceedings of the 40th Annual Conference of the Association for Computer Aided Design in Architecture: Distributed Proximities, ACADIA 2020, 1, pp. 272–281. DOI: https://doi.org/10.52842/conf.acadia.2020.1.272

Du, C., Li, Y., Qiu, Z. y Xu, C. (2023). Stable Diffusion is Unstable. https://github.com/duchengbin8/Stable_Diffusion_is_Unstable

Eisendorf, M. [@marcusisntcool]. (marzo de 2024). Instagram post March 2024 [Video]. Instagram. https://www.instagram.com/reel/C3_Tv-ixmc3/

Fernández-Álvarez, Á. J. y López-Chao, V. (2023). Drawing, scripting, prompting. A critical approach from architectural graphics. Diségno, 13, 143-152. https://doi.org/10.26375/disegno.13.2023.16.13.2023.16

Fraunberger, A. (2023). Linkedin post August 2023. LinkedIn. https://www.linkedin.com/posts/dr-andreas-fraunberger_xr-avatars-genai-ugcPost-7095326558659260416-8a6H/

Game Developers Conference. (2022). Physiological Effects of Crunch: A Look at the Science [Video]. YouTube. https://www.youtube.com/watch?v=Sb2U_9IGgc0

Grepl-Malmgren, L. y Hallenbom, L. (2023). How Does the Level Design in The Last of Us Part 2 Use Lighting and Set Dressing to Suggest Whether an Area Is Safe or Dangerous?

Hu, E., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L. y Chen, W. (2022). Lora: Low-Rank Adaptation of Large Language Models. ICLR 2022 - 10th International Conference on Learning Representations, 1–4. https://doi.org/10.2312/pg.20231276

López-Chao, V., Fernández-Álvarez, Á. J. y Grela, M. R. (2023). Unreal memories: The collective image of architecture in visual social networks. SOBRE: Prácticas Artísticas y Políticas de La Edición, 9(1), 31-42. DOI: https://doi.org/10.30827/sobre.v9i1.27015

Lorenzo-Eiroa, P. y Sprecher, A. (2013). Architecture in Formation: On the Nature of Information in Digital Architecture. https://doi.org/10.4324/9781315890128 DOI: https://doi.org/10.4324/9781315890128

Meira-Rodríguez, P. y López-Chao, V. (2024). Diffusion Models for Environment Visualization: Leveraging Stable Diffusion as a Generator for Architectural Spatial Design. En L. Hermida González, J. P. Xavier, A. Amado Lorenzo y Á. J. Fernández-Álvarez (Eds.), Graphic Horizons. EGA 2024. Springer Series in Design and Innovation (pp. 417-426). Springer. https://doi.org/10.1007/978-3-031-57575-4_49 DOI: https://doi.org/10.1007/978-3-031-57575-4_49

Musli, H. (marzo de 2024). Explorando la intersección de la inteligencia artificial y la arquitectura en Roma [Publicación de estado]. LinkedIn. https://www.linkedin.com/posts/hmusli_aiarchitecture-aiart-rome-ugcPost-7174513123406602240-ceM-

Nuray, Ö. (enero de 2024). LinkedIn post January 2024 [Publicación en LinkedIn]. LinkedIn. https://www.linkedin.com/posts/omer-nuray_install-comfyui-ugcPost-7148360838360342528-fDyg/

Nuray, Ö. (marzo de 2024). ComfyUI for Everything (other than stable diffusion) [Video]. YouTube. https://www.youtube.com/watch?v=fUcDAExxndQ&t=1559s

Petráková, L. y Šimkovič, V. (2023). Architectural alchemy: Leveraging Artificial Intelligence for inspired design–a comprehensive study of creativity, control, and collaboration. Architecture Papers of the Faculty of Architecture and Design STU, 28(4), 3-14. DOI: https://doi.org/10.2478/alfa-2023-0020

Seleit, I. (marzo de 2024). LinkedIn post March 2024. LinkedIn. https://www.linkedin.com/posts/ismailseleit_ai-aiarchitecture-architecture-activity-7169239970254151680-G9KU/?utm_source=share&utm_medium=member_ios

Song, J. y Yip, D. (2023). Exploring the Intersection of AI Art and Film: A Case Study of Giant. Proceedings - 2023 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2023, pp. 347–352. https://doi.org/10.1109/ICMEW59549.2023.00066 DOI: https://doi.org/10.1109/ICMEW59549.2023.00066

Strossmayera, J. J. y Fakultet, O. (2023). Text-to-image Stable Diffusion model. University of Osijek.

Torres-Ferreyros, C. M., Festini-Wendorff, M. A. y Shiguihara-Juarez, P. N. (2017). Developing a videogame using unreal engine based on a four stages methodology. Actas del IEEE ANDESCON 2016. https://doi.org/10.1109/ANDESCON.2016.7836249 DOI: https://doi.org/10.1109/ANDESCON.2016.7836249

Wade, A. C. (2007). The State of the Art: Western Modes of Videogame Production. https://www.researchgate.net/publication/229035979

Williams, G. y Wuetherick, M. (2018). Artist Workflow Improved: Digital Content Creation Tools Roundtripping y Worldbuilding (Presented by Unity Technologies). GDC Vault. https://www.gdcvault.com/play/1024836/Artist-Workflow-Improved-Digital-Content

Zhang, Z., Fort, J. M. y Mateu, L. G. (2023). Exploring the Potential of Artificial Intelligence as a Tool for Architectural Design: A Perception Study Using Gaudí’s Works. Buildings, 13(7), 1863. https://doi.org/10.3390/buildings13071863 DOI: https://doi.org/10.3390/buildings13071863

Žnidarič, M. (enero de 2024). LinkedIn post January 2024. LinkedIn. https://www.linkedin.com/posts/znidaricmarcel_architecture-characterdesign-ai-ugcPost-7147147404998336512-SODW/?utm_source=share&utm_medium=member_ios