TECHNOLOGIES FOR TRANSFORMING 2D IMAGES OF BUILDINGS INTO 3D MODELS USING GENERATIVE ARTIFICIAL INTELLIGENCE ALGORITHMS ON THE EXAMPLE OF THE NEW TASHKENT PROJECT
Keywords:
Artificial intelligence, generative models, neural networks, three-dimensional modeling, digital twins, spatial reconstruction, urban planning information systems.Abstract
In modern urbanization processes, the automation of architectural visualization and the creation of digital twins have become primary factors in developing technological infrastructure. Within the framework of this research, a methodology was developed for automatically converting two-dimensional static images into three-dimensional digital mock-ups using deep learning architectures. The proposed solution is based on the synthesis of neural networks and generative models, allowing for the highly accurate reconstruction of the spatial dimensions and textural characteristics of objects. The dynamics of the obtained results demonstrated a sharp increase in the processing speed of architectural blueprints, alongside the optimization of computational resources for visual data analysis and volumetric modeling algorithms. The conceptual conclusions of the study present a practical platform aimed at reducing design costs and facilitating the formation of a digital urban cadastre through the widespread integration of artificial intelligence into urban planning information systems.
Downloads
References
1. Mildenhall B, Srinivasan PP, Tancik M, Barron JT, Ramamoorthi R, Ng R. NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis. Communications of the ACM. 2021;65(1):99-106.
2. Chen X, Xu L, Yang Y, et al. Deep Learning for 3D Reconstruction: A Survey. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2022;44(12):9889-9907.
3. Wang P, Liu Y, Chen Z, Liu L, Theobalt C. NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction. Advances in Neural Information Processing Systems. 2021;34:27171-27183.
4. Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative Adversarial Networks. Communications of the ACM. 2020;63(11):139-144.
5. Zhang Y, Li S, Zhang X. Digital Twin City: Concepts, Technologies, and Applications. IEEE Internet of Things Journal. 2023;10(5):4512-4525.
6. Kholikov B, Rakhmatov A. Development of Smart City Infrastructure in Uzbekistan: Challenges and Digital Solutions. Asian Journal of Technology & Management. 2022;14(2):112-120.
7. Zhu JY, Park T, Isola P, Efros AA. Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks. IEEE International Conference on Computer Vision (ICCV). 2020;2223-2232.
8. Gao J, Chen W, Xiang T, et al. CityNeRF: Building NeRF at City Scale. IEEE/CVF International Conference on Computer Vision. 2023;14540-14549.
9. Abdullaev U, Karimov R. Geographic Information Systems and 3D Cadastre in Urban Planning of Uzbekistan. Journal of Geographic Information System. 2024;16(1):45-58.
10. O'zbekiston Respublikasi Prezidentining "Yangi Toshkent shahrini barpo etish chora-tadbirlari to'g'risida"gi qarori (PQ-XX). Lex.uz. 2023.
11. Remondino F, Spera MG, Nocerino E, Menna F, Nex F. State of the Art in High-Density Image Matching. The Photogrammetric Record. 2019;29(146):144-166.
12. Liu H, Li Y, Yang J. Hybrid GAN and NeRF Architectures for Complex Architectural Reconstruction. Artificial Intelligence Review. 2024;57(3):88-105.
13. Smith R, Johnson A. Computational Efficiency of Generative AI in Urban Data Processing. Journal of Computing in Civil Engineering. 2025;39(1):04024033.
14. Isroilov MA, Qulmamatov OS. Raqamli shaharsozlikda sun'iy intellekt texnologiyalarini qo'llash istiqbollari. Axborot texnologiyalari va innovatsiyalar jurnali. 2025;7(4):21-29.
15. Tancik M, Casser V, Yan X, et al. Block-NeRF: Scalable Large Scene Neural View Synthesis. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2022;8248-8258.
