CREATING GENERATIVE DESIGNS FOR 3D CHARACTERS BASED ON STRUCTURAL IMAGE ANALYSIS
DOI:
https://doi.org/10.32782/2415-8151.2025.35.39Keywords:
generative model, polygonal mesh, 3D character, component element, computer modeling, single polygon mesh, visible segments, image analysisAbstract
In recent years, generative models for creating 3D objects have shown significant progress, but existing approaches are often limited to creating models represented by a single polygonal mesh. This limits the functionality and adaptability of the created objects. This article discusses the need to develop 3D character generators based on deep generative models that are able to understand and analyze the structure of an object, and identify its components elements and create multigrid views. Unlike existing solutions that create a single polygonal mesh, the proposed approach is aimed at creating models with a clear division into functional components, which significantly increases their practical applicability. The purpose. Create a method that analyzes the structural composition of objects to generate 3D models composed of multiple functionally necessary parts, rather than a single mesh. Methodology. The following methods were used in the study: 1) analytical method by which the literature was analyzed; 2) theoretical and conceptual method, which allowed to determine the conditions necessary for the introduction of IT technology in cultural and artistic practice; The study used methods of computer modeling and analysis, which increased the accuracy of the results. Results. Development of a New Method and successfully proposed an approach using pre-trained models to analyze object ima ges. Scientific novelty. Introduction of a new method for analyzing object structure that goes beyond traditional single-mesh generation and development of an innovative approach that combines pre-trained models for image analysis with segmented 3D model generation Practical relevance. This practical significance demonstrates the research's value across multiple industries and applications, making it particularly relevant for real-world implementation. 1. Simplifies the process of creating game characters 2. Enables dynamic real-time changes to character appearance and structure 3. Improves character animation capabilities through segmented models.
References
Dering M., Cunningham J., Desai R., Yukish M. A., Simpson T. W., Tucker C. S. A Physics-Based Virtual Environment for Enhancing the Quality of Deep Generative Designs. ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, Quebec City, Quebec, Canada, 2018. 54 p .
Fujita K., Minowa K., Nomaguchi Y., Yamasaki S., Yaji K. Design Concept Generation With Variational Deep Embedding Over Comprehensive Optimization. International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, Virtual, 2021. 38 p.
Furferi R. Deep Learning Approaches for 3D Model Generation from 2D Artworks to Aid Blind People with Tactile Exploration. 2024. 12 p.
Girdhar R., Fouhey D. F., Rodriguez M., Gupta A. Learning a Predictable and Generative Vector Representation for Objects. European Conference on Computer Vision, Springer, Amsterdam, 2016. Pp. 484–499.
Goodfellow I., Pouget-Abadie J., Mirza M., Xu B., Warde-Farley D., Ozair S., Courville A., Bengio Y. Generative Adversarial Nets. Advances in Neural Information Processing Systems, Palais des Congrès de Montréal, Montréal, 2014. Pp. 2672–2680.
Guillard B., Remelli E., Yvernay P., Fua P. Sketch2mesh: Reconstructing and Editing 3D Shapes From Sketches. IEEE/CVF International Conference on Computer Vision, Virtual, 2021. 10 p.
Gunpinar E., Ovur S. E., Gunpinar S. A User-Centered Side Silhouette Generation System for Sedan Cars Based on Shape Templates. Optim. Eng., 2019. Vol. 20(3). Pp. 683–723.
Han X., Gao C., Yu Y. Deepsketch2face: A Deep Learning Based Sketching System for 3D Face and Caricature Modeling. ACM Trans. Graph., 2017. Vol. 36(4). Pp. 1–23.
Jin A., Fu Q., Deng Z. Contour-Based 3D Modeling Through Joint Embedding of Shapes and Contours. Symposium on Interactive 3D Graphics and Games, San Francisco, CA, 2020. Pp. 1–10.
Kingma D. P., Welling M. Auto-Encoding Variational Bayes. Preprint arXiv, 2013. 114 p.
Lun Z., Gadelha M., Kalogerakis E., Maji S., Wang R. 3D Shape Reconstruction From Sketches Via Multi-View Convolutional Networks. 2017 International Conference on 3D Vision (3DV), IEEE, Qingdao, 2017. Pp. 67–77.
Nozawa N., Shum H. P., Feng Q., Ho E. S., Morishima S. 3D Car Shape Reconstruction From a Contour Sketch Using GAN and Lazy Learning. Vis. Comput., 2021. Vol. 38(4). Pp. 1317–1330.
Nozawa N., Shum H. P., Ho E. S., Morishima S. Single Sketch Image Based 3D Car Shape Reconstruction With Deep Learning and Lazy Learning. 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, Valetta, 2020. Pp. 179–190.
Oh S., Jung Y., Kim S., Lee I., Kang N. Deep Generative Design: Integration of Topology Optimization and Generative Models. ASME J. Mech. Des., 2019. Vol. 141(11). 405 p.
Pratt M. J., Anderson B. D., Ranger T. Towards the Standardized Exchange of Parameterized Feature-Based CAD Models. Comput. Aided Des., 2005. Vol. 37(12). Pp. 1251–1265.
Reid T. N., Gonzalez R. D., Papalambros P. Y. Quantification of Perceived Environmental Friendliness for Vehicle Silhouette Design. ASME J. Mech. Des., 2010. Vol. 132(10). 1010 p.
Shu D., Cunningham J., Stump G., Miller S. W., Yukish M. A., Simpson T. W., Tucker C. S. 3D Design Using Generative Adversarial Networks and Physics-Based Validation. ASME J. Mech. Des., 2020. Vol. 142(7). 1701 p.
Xiang N., Wang R., Jiang T., Wang L., Li Y., Yang X., Zhang J. Sketch-Based Modeling With a Differentiable Renderer. Comput. Animat. Virtual Worlds, 2020. Vol. 31(4–5). e1939.
Yuan Y.-J., Lai Y.-K., Yang J., Duan Q., Fu H., Gao L. Mesh Variational Autoencoders With Edge Contraction Pooling. IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, Virtual, 2020. Pp. 274–275.
Zhang W., Yang Z., Jiang H., Nigam S., Yamakawa S., Furuhata T., Shimada K., Kara L. B. 3D Shape Synthesis for Conceptual Design and Optimization Using Variational Autoencoders. International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, Anaheim, CA, 2019. P. V02AT03A.
