A StyleGAN2-based framework for generating and evaluating urban color landscapes and street vitality

Abstract

Urban color landscapes play a significant role in shaping perceptual experience and street vitality. We proposed a style-based generative adversarial network 2 (StyleGAN2)-inspired generative framework for creating urban colorscapes and quantitatively assessing their vitality. The approach integrated advanced data preprocessing, generator–discriminator architecture, and hyperparameter optimization using a non-saturating logistic loss function. Vitality was evaluated through three chromatic indicators-saturation, contrast, and diversity-and validated against behavioral (pedestrian volume) and socioeconomic (point-of-interest density) data via correlation analysis (r = 0.47–0.68). The model achieved theoretical convergence (Fréchet Inception Distance < 15) and optimality, while ablation experiments with a deep convolutional GAN, Wasserstein GAN with gradient penalty, and StyleGAN3 confirmed its superior generative performance. The synthesized images exhibited an 18.2% increase in saturation and a 10.5% increase in diversity relative to real-world scenes, suggesting a strong positive association with urban vitality, as established in our correlation analysis. Computational efficiency was enhanced through mixed-precision training, reducing total processing time. Empirical and perceptual validations confirmed the framework’s robustness, offering a reproducible pathway for artificial intelligence-driven urban color planning.

Keywords

Adversarial networks / StyleGAN2 / Urban colorscape generative / Visual analytics / Vitality assessment

References

[1] Karras T, Laine S, Aila T. A style-based generator architecture for generative adversarial networks. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2019: 4396-4405. https://doi.org/10.1109/CVPR.2019.00453

[2] Batty M. The New Science of Cities. Cambridge, MA, USA: MIT Press; 2017.

[3] Karras T, Aittala M, Hellsten J, Laine S, Lehtinen J, Aila T. Training generative adversarial networks with limited data. In: Advances in Neural Information Processing Systems. Curran Associates, Inc. 2020: 12104-12114. https://papers.nips.cc/paper/2020/hash/8d30aa96e72440759f74bd2306c1fa3d-Abstract.html. Accessed November 1, 2025.

[4] Jacobs J. The Death and Life of Great American Cities. Vintage Books; 1961.

[5] Chen R, Zhao J, Yao X, et al. Enhancing urban landscape design: a GAN-based approach for rapid color rendering of park sketches. Land. 2024; 13(2): 254. https://doi.org/10.3390/land13020254

[6] Remtulla R, Samet A, Kulbay M, et al. A future picture: a review of current generative adversarial neural networks in vitreoretinal pathologies and their future potentials. Biomedicines. 2025; 13(2): 284. https://doi.org/10.3390/biomedicines13020284

[7] Farooq MA, Yao W, Costache G, Corcoran P. ChildGAN: large scale synthetic child facial data using domain adaptation in StyleGAN. IEEE Access. 2023; 11: 108775-108791. https://doi.org/10.1109/ACCESS.2023.3321149

[8] Karras T, Laine S, Aittala M, Hellsten J, Lehtinen J, Aila T. Analyzing and improving the image quality of StyleGAN. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2020: 8107-8116. https://doi.org/10.1109/CVPR42600.2020.00813.

[9] Gan W, Zhao Z, Wang Y, Zou Y, Zhou S, Wu Z. UDGAN: a new urban design inspiration approach driven by using generative adversarial networks. J Comput Des Eng. 2023; 11(1): 305-324. https://doi.org/10.1093/jcde/qwae014

[10] Meng S. Exploring in the latent space of design: a method of plausible building facades images generation, properties control and model explanation base on StyleGAN2. In: Yuan PF, Chai H, Yan C, Leach N, eds. Proceedings of the 2021 DigitalFUTURES. Singapore: Springer; 2022: 55-68. https://doi.org/10.1007/978-981-16-5983-6_6.

[11] Jaglarz A. Perception of color in architecture and urban space. Buildings. 2023; 13: 2000. https://doi.org/10.3390/buildings13082000

[12] Zhang A, Li W, Wu J, Lin J, Chu J, Xia C. How can the urban landscape affect urban vitality at the street block level? A case study of 15 metropolises in China. Environ Plan B Urban Anal City Sci. 2021; 48(5): 1245-1262. https://doi.org/10.1177/2399808320924425

[13] Choi J, Seo K, Ashtari A, Noh J. StyleCineGAN: landscape cinemagraph generation using a pre-trained StyleGAN. In: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA, USA: IEEE; 2024: 7872-7881. https://doi.org/10.1109/CVPR52733.2024.00752

[14] Nami P, Jahanbakhsh P, Fathalipour A. The role and heterogeneity of visual pollution on the quality of urban landscape using GIS; case study: historical garden in city of Maraqeh. Open J Geol. 2016; 6: 20-29. https://doi.org/10.4236/ojg.2016.61003

[15] Li X, Xue S, Li Z, Fang X, Zhu T, Ni C. A candy defect detection method based on StyleGAN2 and improved YOLOv7 for imbalanced data. Foods. 2024; 13(20): 3343. https://doi.org/10.3390/foods13203343

[16] Schaerf L, Alfarano A, Postma E. ColorwAI: Generative Colorways of Textiles Through GAN and Diffusion Disentanglement. 2024. https://doi.org/10.48550/arXiv.2407.11514

[17] Way D-L, Lo C-H, Wei Y-H, Shih Z-C. TwinGAN: twin generative adversarial network for Chinese landscape painting style transfer. IEEE Access. 2023; 11: 60844-60852. https://doi.org/10.1109/ACCESS.2023.3274666

[18] Deng B. Style transfer for converting images into Chinese landscape painting based on CycleGAN. Appl Comput Eng. 2024; 51: 53-58. https://doi.org/10.54254/2755-2721/51/20241162

[19] Lee M, Lee H, Hwang S, Choi M. Understanding the impact of the walking environment on pedestrian perception and comprehension of the situation. J Transp Health. 2021; 23: 101267. https://doi.org/10.1016/j.jth.2021.101267

[20] Fedorova S. GANs for Urban Design. 2021. https://doi.org/10.48550/arXiv.2105.01727

[21] Sun Q, Chen Y, Tao W, et al. A GAN-based approach toward architectural line drawing colorization prototyping. Vis Comput. 2022; 38: 1283-1300. https://doi.org/10.1007/s00371-021-02219-x

[22] Tian R. Suggestive site planning with conditional GAN and urban GIS data. In: Proceedings of the 2020 DigitalFutures. Singapore: Springer; 2021: 103-113. https://doi.org/10.1007/978-981-33-4400-6_10

[23] Chang K-H, Cheng C-Y, Luo J, Murata S, Nourbakhsh M, Tsuji Y. Building-GAN: Graph-conditioned Architectural Volumetric Design Generation. 2021. https://doi.org/10.48550/arXiv.2104.13316

[24] Sun C, Zhou Y, Han Y. Automatic generation of architecture facade for historical urban renovation using generative adversarial network. Build Environ. 2022; 212: 108781. https://doi.org/10.1016/j.buildenv.2022.108781

[25] Wu Z, Lischinski D, Shechtman E. StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation. 2020. https://doi.org/10.48550/arXiv.2011.12799

[26] Zhang H, Goodfellow I, Metaxas D, Odena A. Self-attention generative adversarial networks. In: Proceedings of the 36th International Conference on Machine Learning PMLR. 2019: 7354-7363. https://proceedings.mlr.press/v97/zhang19d.html. Accessed November 1, 2025.

[27] Sauer A, Schwarz K, Geiger A. StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets. 2022. https://doi.org/10.48550/arXiv.2202.00273

[28] Xu S, Jiang H, Wang H. A generative urban form design framework based on deep convolutional GANs and landscape pattern metrics for sustainable renewal in highly urbanized cities. Sustainability. 2025; 17(10): 4548. https://doi.org/10.3390/su17104548

[29] Zhang R, Tang X, Wu L, Wang Y, He X, Liu M. Evaluation on AI-generative emotional design approach for urban vitality spaces: a LoRA-driven framework and empirical research. Land. 2025; 14(6): 1300. https://doi.org/10.3390/land14061300

[30] Yu M, Zheng X, Qin P, Cui W, Ji Q. Urban color perception and sentiment analysis based on deep learning and street view big data. Appl Sci. 2024; 14(20): 9521. https://doi.org/10.3390/app14209521

[31] Zhang X, Han H, Qiao L, et al. Emotional-health-oriented urban design: a novel collaborative deep learning framework for real-time landscape assessment by integrating facial expression recognition and pixel-level semantic segmentation. Int J Environ Res Public Health. 2022; 19(20): 13308. https://doi.org/10.3390/ijerph192013308

[32] Zhang C, Chen Y, Dewancker BJ, et al. Emotional landscapes in urban design: analyzing color emotional responses of the elderly to community outdoor spaces in Yi Jie Qu. Buildings. 2024; 14(13): 793. https://doi.org/10.3390/buildings14030793

[33] Huang J, Johanes M, Kim FC, Doumpioti C, Holz G-C. On GANs, NLP and architecture: combining human and machine intelligences for the generation and evaluation of meaningful designs. Technol Archit Des. 2021; 5(2): 207-224. https://doi.org/10.1080/24751448.2021.1967060

[34] Xu H, Omitaomu O, Sabri S, Zlatanova S, Li X, Song Y. Leveraging Generative AI for Smart City Digital Twins: a Survey on the Autonomous Generation of Data, Scenarios, 3D City Models, and Urban Designs. 2024. https://doi.org/10.48550/arXiv.2405.19464

[35] He M, Liang Y, Wang S, et al. Generative AI for Urban Design: a Stepwise Approach Integrating Human Expertise with Multimodal Diffusion Models. 2025. https://doi.org/10.48550/arXiv.2505.24260

[36] Jiang Y, Zhang Y, Liu Y, Huang Z. A review of urban vitality research in the Chinese world. Trans Urban Data Sci Technol. 2023; 2(2-3): 81-99. https://doi.org/10.1177/27541231231154705

[37] Wang Z, Xia N, Zhao X, Gao X, Zhuang S, Li M. Evaluating urban vitality of street blocks based on multi-source geographic big data: a case study of Shenzhen. Int J Environ Res Public Health. 2023; 20(5): 3821. https://doi.org/10.3390/ijerph20053821

[38] Tang Q, Zheng L, Chen Y, Yan L, Chen J. Artificial intelligence empowering museum space layout design: insights from China. PLoS One. 2024; 19(11): e0310594. https://doi.org/10.1371/journal.pone.0310594

[39] Eskandar G, Farag Y, Yenamandra T, Cremers D, Guirguis K, Bin Y. Urban-StyleGAN: Learning to Generate and Manipulate Images of Urban Scenes. 2023. https://doi.org/10.48550/arXiv.2305.09602

[40] Lee G, Yim J, Kim C, Kim M. StyLandGAN: a StyleGAN Based Landscape Image Synthesis Using Depth-map. 2022. https://doi.org/10.48550/ARXIV.2205.06611

[41] Elliot AJ, Maier MA. Color psychology: effects of perceiving color on psychological functioning in humans. Annu Rev Psychol. 2014; 65(1): 95-120. https://doi.org/10.1146/annurev-psych-010213-115035

[42] Li C, Zhang T, Du X, Zhang Y, Xie H. Generative AI models for different steps in architectural design: a literature review. Front Archit Res. 2024; 14(3): 759-783. https://doi.org/10.1016/j.foar.2024.10.001

[43] Sauer A, Karras T, Laine S, Geiger A, Aila T. StyleGAN-T: unlocking the power of gans for fast large-scale text-to-image synthesis. In: Proceedings of the 40th International Conference on Machine Learning PMLR. 2023: 30105-30118. https://proceedings.mlr.press/v202/sauer23a.html. Accessed November 1, 2025.

[44] Ali M, Rossi L, Bertozzi M. CoLoR-GAN: Continual Few-shot Learning with Low-Rank Adaptation in Generative Adversarial Networks. 2025. https://doi.org/10.48550/arXiv.2510.13869

[45] Qian B. Intelligent interior design based on BachGAN and StyleGAN2. In: EEI 2022 4th International Conference on Electronic Engineering and Informatics. 2022: 1-7. https://ieeexplore.ieee.org/document/10048242. Accessed November 1, 2025.

[46] Wang Z, Shen M, Huang Y. Combining eye-tracking technology and subjective evaluation to determine building facade color combinations and visual quality. Appl Sci. 2024; 14(18): 8227. https://doi.org/10.3390/app14188227

[47] Wang Z, Shen M, Huang Y. Exploring the impact of facade color elements on visual comfort in old residential buildings in Shanghai: insights from eye-tracking technology. Buildings. 2024; 14(6): 1758. https://doi.org/10.3390/buildings14061758

[48] Borji A. Pros and cons of GAN evaluation measures. Comput Vis Image Underst. 2019; 179: 41-65. https://doi.org/10.1016/j.cviu.2018.10.009

[49] Kynkäänniemi T, Karras T, Laine S, Lehtinen J, Aila T. Improved precision and recall metric for assessing generative models. In: Advances in Neural Information Processing Systems. Curran Associates, Inc. 2019: 3927-3936. https://proceedings.neurips.cc/paper/2019/hash/0234c510bc6d908b28c70ff313743079-Abstract.html. Accessed November 30, 2025.

[50] Chong MJ, Lee H-Y, Forsyth D. StyleGAN of all Trades: Image Manipulation with only Pretrained StyleGAN. 2021. https://doi.org/10.48550/arXiv.2111.01619

[51] Cordts M, Omran M, Ramos S, et al. The cityscapes dataset for semantic urban scene understanding. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016: 3213-3223. https://doi.org/10.1109/CVPR.2016.350

[52] Valdez P, Mehrabian A. Effects of color on emotions. J Exp Psychol Gen. 1994; 123(4): 394-409. https://doi.org/10.1037/0096-3445.123.4.394

[53] Gao X, Xin JH. Investigation of human’s emotional responses on colors. Color Res Appl. 2006; 31(5): 411-417. https://doi.org/10.1002/col.20246

[54] Ou L, Luo MR, Woodcock A, Wright A. A study of colour emotion and colour preference. Part I: colour emotions for single colours. Color Res Appl. 2004; 29(3): 232-240. https://doi.org/10.1002/col.20010

[55] Montgomery J. Making a city: urbanity, vitality and urban design. J Urban Des. 1998; 3(1): 93-116. https://doi.org/10.1080/13574809808724418

[56] Mehta V. Evaluating public space. J Urban Des. 2014; 19(1): 53-88. https://doi.org/10.1080/13574809.2013.854698