Matrix-based character design systems: Parallel principles in manga matrix and lora for neural networks

Anthoni Reza Pahlevi; Alvanov Zpalanzani Mansoor

doi:10.38010/deskomvis.v6i1.96

Authors

Anthoni Reza Pahlevi Bandung Institute of Technology
Alvanov Zpalanzani Mansoor Bandung Institute of Technology

DOI:

https://doi.org/10.38010/deskomvis.v6i1.96

Keywords:

Artificial Intelligence, Character Design, Manga Matrix, Low-Rank Adaptation

Abstract

The integration of AI in character design lacks systematic frameworks that translate traditional visual design principles into computational parameters, creating a conceptual gap between designers' artistic intent and AI tool configuration. This research identifies four parallel principles—structural decomposition, modular recombination, granular control, and hierarchical organization—shared between Hiroyoshi Tsukamoto's Manga Matrix and Low-Rank Adaptation (LoRA), developing the first translation framework that maps visual design decisions to specific LoRA parameters. We reveal how matrix-based organizational principles create actionable pathways for designer-AI collaboration, establishing that these systems share fundamental operational mechanisms despite their distinct domains. Through literature review and comparative analysis, we examine how Manga Matrix's visual grid elements correspond to LoRA's rank values and weight distributions, establishing practical implementation guidelines for each principle. The research establishes foundational methodology for human-AI co-creation, contributing to more practical and consistent AI-assisted design workflows.

Downloads

Download data is not yet available.

References

Avrahami, O., Hertz, A., Vinker, Y., Arar, M., Fruchter, S., Fried, O., Cohen-Or, D., & Lischinski, D. (2023). The Chosen One: Consistent Characters in Text-to-Image Diffusion Models. ArXiv, abs/2311.10093.

Choi, D., Im, J., & Sung, Y. (2024). LoRA Fusion: Enhancing Image Generation. Mathematics, 12(22), 3474. https://doi.org/10.3390/math12223474.

Civitai. (2025). Using Civitai – The On-Site LoRA Trainer. Retrieved on January 12, 2025.

Epstein, Z., Hertzmann, A., Herman, L., Mahari, R., Frank, M., Groh, M., Schroeder, H., Smith, A., Akten, M., Fjeld, J., Farid, H., Leach, N., Pentland, A., & Russakovsky, O. (2023). Art and the science of generative AI. Science, 380, 1110 - 1111. https://doi.org/10.1126/science.adh4451.

Hu, J.E., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., & Chen, W. (2021). LoRA: Low-Rank Adaptation of Large Language Models. ArXiv, abs/2106.09685.

Jones, M., Wang, S. Y., Kumari, N., Bau, D., & Zhu, J. Y. (2024, December). Customizing text-to-image models with a single image pair. In SIGGRAPH Asia 2024 Conference Papers (pp. 1-13).

Kachalin, P. (2025). 25 Famous Memes Reimagined As Art By Studio Ghibli. Retrieved on June 7, 2025

Karaalioğlu, S., & Sayin, Z. (2022). Investigation of character designs in context of visual impact item. The Turkish Online Journal of Design Art and Communication. https://doi.org/10.7456/11203100/020

Kulkarni, C., Druga, S., Chang, M., Fiannaca, A.J., Cai, C.J., & Terry, M. (2023). A Word is Worth a Thousand Pictures: Prompts as AI Design Material. ArXiv, abs/2303.12647.

Lin, S., Wang, A., & Yang, X. (2024). SDXL-Lightning: Progressive Adversarial Diffusion Distillation. ArXiv, abs/2402.13929.

Maliyyaa, H., Fami, A., Paramita, A. G., & Valentina, S. D. (2024). Analisis Desain Karakter dalam Ebook Desa Sambeng Menggunakan Metode Manga Matrix. DeKaVe, 17(1), 61-76

Mateu-Mestre, M. (2010). Framed ink: Drawing and composition for visual storytellers. Design Studio Press.

Miketić, N., Pinćjer, I., & Lilić, A. (2018). Integration of the visual elements of art and personality factors in process of character design. Proceedings of the 9th International Symposium on Graphic Engineering and Design.

Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022). High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 10684–10695).

Smith, J., Hsu, Y., Zhang, L., Hua, T., Kira, Z., Shen, Y., & Jin, H. (2023). Continual Diffusion: Continual Customization of Text-to-Image Diffusion with C-LoRA. ArXiv, abs/2304.06027. https://doi.org/10.48550/arXiv.2304.06027.

Sullivan, K., Schumer, G., & Alexander, K. (2008). Ideas for the Animated Short with DVD: Finding and Building Stories (1st ed.). Routledge. https://doi.org/10.4324/9780080560885

Tillman, B. (2011). Creative Character Design. Focal Press.

Tsai, F. (2007). Effective character design. ImagineFX, 15, 80–83.

Tsukamoto, H. (2006). Manga Matrix: Create unique characters using the Japanese matrix system (p. 172). Collins Design.

Ursyn, A. (2018). Character Development for Visual Storytelling in Traditional and Digital Media. , 304-348. https://doi.org/10.4018/978-1-5225-5332-8.CH013

Xie, Y., Pan, Z., Ma, J., Luo, J., & Mei, Q. (2023). A Prompt Log Analysis of Text-to-Image Generation Systems. Proceedings of the ACM Web Conference 2023.

Zhao, J., Wang, T., Abid, W., Angus, G., Garg, A., Kinnison, J., ... & Rishi, D. (2024). Lora land: 310 fine-tuned llms that rival gpt-4, A technical report. arXiv preprint arXiv:2405.00732.

Zhong, M., Shen, Y., Wang, S., Lu, Y., Jiao, Y., Ouyang, S., Yu, D., Han, J., & Chen, W. (2024). Multi-LoRA Composition for Image Generation.