From Cards to Creatures: Interactive 3D Generation Using GenAI

• Edward Scott Carvalho Johnson AKCIT / UFG
• Arthur Galdino Dangoni AKCIT / UFG
• Pedro Elias Engelberg Silva Borges AKCIT / UFG
• Diogo Fernandes Costa Silva AKCIT / UFG
• Arlindo Rodrigues Galvão Filho AKCIT / UFG

Resumo

We present “From Cards to Creatures”, a mobile augmented reality system that enables interactive 3D visualization of physical trading cards through generative AI. The proposed pipeline combines object detection, retrieval, and image-to-3D generation to support real-time rendering anchored to physical inputs. Specifically, a YOLOv11-based detector identifies and extracts the card’s unique ID from the mobile camera feed, which allows for instant retrieval of the corresponding 3D asset from a pre-compiled dataset. Each card’s artwork is processed with the Hunyuan3D-2.1 generative model to produce high-quality meshes and textures. The retrieved model is rendered and spatially aligned with the detected card in the AR scene. By decoupling the generation process from real-time image capture, the system ensures consistent reconstruction and low-latency inference. A qualitative evaluation using collectible Pokémon cards demonstrates accurate alignment, coherent shape generation, and responsiveness during interaction. This case study illustrates a modular architecture that can be extended to other domains involving physical artifacts and generative 3D content.

Palavras-chave: augmented reality, generative AI, image-to-3D, trading card games

Referências

Milgram, P. Takemura, H. Utsumi, A. and Kishino, F. Augmented reality a class of displays on the reality-virtuality continuum. Telemanipulator and Telepresence Technologies, International Society for Optics and Photonics, 2351:282–292, 1995.

Azuma, R. T. A survey of augmented reality. Presence Teleoperators and Virtual Environments, 6(4):355–385, 1997.

Nguyen, M. Yeap, W. and Hooper, S. Design of a new trading card for table-top augmented reality game environment. 2016 International Conference on Image and Vision Computing New Zealand (IVCNZ), IEEE, Palmerston North, New Zealand, 1–6, 2016.

Bedoya-Rodriguez, S. Gomez-Urbano, C. Uribe-Quevedo, A. and Quintero, C. Augmented reality RPG card-based game. 2014 IEEE Games Media Entertainment (GEM), Toronto, Canada, IEEE, 1–4, 2014.

Karkera, A. Dhadse, S. Gawde, V. and Jain, K. Pokemon fight augmented reality game. 2018 Second International Conference on Inventive Communication and Computational Technologies (ICICCT), IEEE, Coimbatore, India, 1762–1764, 2018.

Team, T. H. Hunyuan3D 2.1 From images to high-fidelity 3D assets with production-ready PBR material, 2025.

Liu, Y. Si, Y. Lc, R. and Harteveld, C. cARd Mixed Reality Approach for a Total Immersive Analog Game Experience. Springer Nature Switzerland AG, 889–906, 2021.

Zuo, T. Hu, J. Spek, E. and Birk, M. Understanding the effect of fantasy in augmented reality game-based learning from a player journey perspective. HCHI 23 Proceedings of the Eleventh International Symposium of Chinese CHI, 55–60, 2024.

Aviortheking. TCGdex The Pokemon TCG Card Database. GitHub Repository, 2020. Available [link]

Kim, T. Kim, K. Lee, J. Cha, D. Lee, J. and Kim, D. Revisiting image pyramid structure for high resolution salient object detection. Proceedings of the Asian Conference on Computer Vision, 108–124, 2022.

Jocher, G. and Qiu, J. Ultralytics YOLO11. GitHub Repository, 2024. Available [link]

Authors, P. PaddleOCR Awesome Multilingual OCR Toolkits based on PaddlePaddle. GitHub Repository, 2020. Available [link]