GenAI ConceptLab: A Generative AI tool for Assisting Artists in Crafting Anime Characters and Narratives

• Matheus Lobo CESAR School
• Pamela T. L. Bezerra CESAR School
• Gabriella A. B. Barros Profissional Independente

DOI: https://doi.org/10.5753/sbgames.2025.10147

Resumo

Introduction: Generative AI has impacted how artists, either professional or amateur, create art. However, many generative tools require extensive training to develop good prompts and select the appropriate model features to produce high-quality results. Objective: To reduce the necessary learning curve and technical understanding of generative AI, this paper proposes a web application that integrates generative image and text models to create anime character concept art and minibiographies. The main goal is to assist independent creators and anime enthusiasts by providing an accessible and scalable tool that minimizes prompt engineering through an intuitive interface. Methodology or Steps: To develop the proposed solution, an initial study of similar solutions and generative models focused on anime style was conducted. This was followed by the development of the web application and the exploration of different prompts and interfaces. Finally, system and user evaluations were performed to analyze the solution’s capabilities in generating images and text and to collect users’ opinions. Results: User tests showed positive feedback on the interface and the quality of content generated. Meanwhile, the system’s tests indicated that the solution is capable of generating art for a wide range of anime genres, with a positive correlation between diversity and the amount of details provided.

Palavras-chave: Generative AI Models, Concept Art, Anime Style, Game Art

Referências

Esser, P., Kulal, S., Blattmann, A., Entezari, R., Müller, J., Saini, H., Levi, Y., Lorenz, D., Sauer, A., Boesel, F., et al. (2024). Scaling rectified flow transformers for high-resolution image synthesis. In Forty-first international conference on machine learning.

Grand View Research (2022). Anime market analysis by product type (tv shows, movies, merchandising, others), by revenue (usd billion), and segment forecasts, 2025 - 2032. [link]. Accessed: 2025-02-28.

Holara (2023). Holara. [link].

Hollywood Reporter (2022). Japanese anime: The world’s most bankable genre. [link] Accessed: 2025-02-28.

Intelligent Systems (1990). Fire Emblem. Digital.

Midjourney, I. (2022). Midjourney. [link].

Monolith Soft (2010). Xenoblade. Digital.

OpenAI (2021). Dall-e: Creating images from text. [link].

OpenAI (2023). ChatGPT: OpenAI Language Model. [link].

Rombach, R., Blattmann, A., Lorenz, D., Esser, P., e Ommer, B. (2022). High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695.

Sizigi Studios (2019). Waifu Labs. [link].

Spellbrush (2022). niji journey. [link].

Spotlight Report (2024). The rising popularity of anime-themed games.

Square Enix (1987). Fire emblem. Digital.