Utilizing Machine Learning in the Architectural Design Process: Exploration of Generating Diverse Spatial Layout Designs Using Generative Adversarial Networks

Subject Areas :

Mahsa Hamouni ¹ , Hossein Soltanzadeh ^{2 *} , Hadi ghoddusifar ³ , Muharram Mansoorizadeh ⁴

1 - Department of Architecture, Faculty of Art & Architecture, South Tehran Branch, Islamic Azad University, Tehran, Iran
2 - Department of Architecture and urban planning, Faculty of Architecture and urban planning , Central Tehran Branch, Islamic Azad University ,Tehran ,Iran.
3 - Assistant Professor of Architecture, Faculty of Art and Architecture, South Tehran Branch, Islamic Azad University, Tehran, Iran
4 - Associate Professor, Department of Computer Engineering, Bu-Ali Sina University, Hamadan, Iran

Received: 2024-07-09 Accepted : 2024-10-23 Published : 2025-06-17

Keywords: Artificial intelligence, Generative adversarial network, space layout, bubble diagram,

Abstract :

Introduction: Artificial intelligence technology has become a trending topic in the field of architectural layout design. The core technology of artificial intelligence, machine learning, has attracted the attention of architects. Numerous design-related disciplines could be impacted by the capacity of this emerging field to learn from examples and extrapolate that knowledge into the creation of new instances. As a machine learning model, the generative adversarial network (GAN) has demonstrated remarkable outcomes in the development of spatial layouts within defined boundaries. However, these approaches generate a single output for each input condition. This paper has a new approach to generate a variety of space layout designs with more than one output for a given boundary. The main idea of this study is to control the generated results by using the bubble diagram so that we can have a diversity layout according to the same boundary conditions. Methodology: For this purpose, a specific dataset is prepared. Then a directory holding the bubble diagram images that serve as extra conditions is loaded by modifying the data loader. The two input tensors from an image pair consisting of an input and a condition image are concatenated along the C dimension. Results: We test the model by using the test set. The outputs of the model are evaluated based on quantitative and qualitative methods. The results show that the generated layouts are relatively ideal. The bubble diagram provides users with information about what is included in the synthetic space layout plans, and the generated plans also adhere to the boundaries of buildings. Conclusion: This research enables designers to control the results and participate in the process of machine learning generative design. Within the same building boundaries, users can generate different floor plan layouts by giving various bubble diagrams.

References:


• As, I., Pal, S., & Basu, P. (2018). Artificial intelligence in architecture: Generating conceptual design via deep learning. International Journal of Architectural Computing. https:// doi.org /10.1177/14780771188 00 982.
• Caetano, I., Santos, L., & Leitão, A. (2020). Computational design in architecture: Defining parametric, generative, and algorithmic design. Frontiers of Architectural Research, 9(2), 287-300. DOI:10.1016/j.foar.2019.12.008.
• Chaillou, S. (2019). Architecture, towards a new approach, Harvard. https://doi.org/10.9783/9781949057027-006.

• Creswell, A., T. White, V. Dumoulin, K. Arulkumaran, B. Sengupta, A., & Bharath. A. (2018). “Generative Adversarial Networks: An Overview.” IEEE Signal Processing Magazine 35, no. 1: 53–65. https:// doi.org/10.1109/MSP.2017.2765202.

• Cross, N., (2011). Design thinking: understanding how designers think and work. Oxford: Bloomsbury Academic.

• Goodfellow, I. J., J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde- Farley, S. Ozair, A. Courville, and Y. Bengio. 2014. “Generative Adversarial Nets.” In Proceedings of the Twenty-Seventh International Conference on Neural Information Processing Systems (NIPS), 2:2672–2680. Montreal, Canada, December 8–13.

• Goodfellow, I., Y. Bengio, and A. Courville. 2016. Deep Learning. Cambridge, MA: MIT Press. https://doi.org/10.1016/B978-0-12- 391420-0.09987-X. ISBN 9780262035613.

• Grover, R., Emmitt, S. & Copping, A. (2018). The typological learning framework: the application of structured precedent design knowledge in the architectural design studio. International Journal of Technology and Design Education. 28. 10.1007/s10798-017-9421-4.
• Hong, T., Wang, Z., Luo, X., & Zhang, W. (2020). State-of-the-Art on Research and Applications of Machine Learning in the Building Life Cycle. Energy and Buildings. 212. 10.1016/j.enbuild.2020.109831.

• Isola, Ph., Zhu, J-Y., Zhou, T., & Efros, A. (2017). Image-to-Image Translation with Conditional Adversarial Networks. 5967-5976.10.11 09/CVPR.2017.632.
• Kanters, J., Horvat, M., (2012). The Design Process known as IDP: A Discussion, Energy Procedia, Volume 30, 2012, Pages 1153-1162, ISSN 1876-6102, https://doi.org/10.1016/j.egypro.2012.11.128.
• Karadag, I., Zeynel Güzelci, O., & Alaçam, S. (2022). "EDU-AI: a twofold machine learning model to support classroom layout generation", Construction Innovation. 10.1108/CI-02-2022-0034.

• Lawson, B., & Dorst, K. (2009). Design Expertise (1st ed.). Routledge. https://doi.org/10.4324/9781315072043.
• Liu, Y., Fang, Ch., Yang, Z., Wang, X., Zhou, Z., Deng, Q. & Liang, L. (2022). Exploration on Machine Learning Layout Generation of Chinese Private Garden in Southern Yangtze. In: Yuan, P.F., Chai, H., Yan, C., Leach, N. (eds) Proceedings of the 2021 DigitalFUTURES. CDRF 2021. Springer, Singapore. https://doi.org/10.1007/978-981-16-5983-6_4.

• Liu, Y., Fang, Ch., Yang, Z., Wang, X., Zhou, Z., Deng, Q. & Liang, L. (2022). Exploration on Machine Learning Layout Generation of Chinese Private Garden in Southern Yangtze. In: Yuan, P.F., Chai, H., Yan, C., Leach, N. (eds) Proceedings of the 2021 DigitalFUTURES. CDRF 2021. Springer, Singapore. https://doi.org/10.1007/978-981-16-5983-6_4.

• Lu, Stephen & Liu, Ang. (2011). Subjectivity and objectivity in design decisions. Cirp Annals-manufacturing Technology - CIRP ANN-MANUF TECHNOL. 60. 161-164. 10.1016/j.cirp.2011.03.122.

• Newton, D. (2019). Generative Deep Learning in Architectural Design. Technology|Architecture + Design. 3. Doi: 176-189. 10 .1080/ 2475 1448. 2019. 1 6 40536.

• Özerol, G., & Arslan Selçuk, S. (2022). Machine learning in the discipline of architecture: A review on the research trends between 2014 and 2020. International Journal of Architectural Computing. https://doi.org/10.1177_14780771221100102.
• Purcell, T., & Gero, J. (1998). Drawings and the design process. Design Studies. DOI 19. 10.1016/S0142-694X(98)00015-5.
• Russell, S. J. and P. Norvig. 2016. Artificial Intelligence: A Modern Approach. 3rd ed. Boston: Pearson Education.

• Visser W., (2006). The cognitive artifacts of designing. Hillsdale, NJ: Lawrence Erlbaum Associates. Wang X, Yang Y. & Zhang K. Customization and generation of floor plans based on graph transformations. Automation in Construction. (2018); DOI: 94. 405-416. 10.1016/j.autcon.2018.07.017.

• Wang, T.-C., Liu, Y., Zhu, J., Tao, A., Kautz, J., & Catanzaro, B. (2018). “High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Doi: 10.1109/CVPR.2018.00917.https://arxiv.org/ abs/17 1 1.11585.

• Yang, Z.; Gan, Z.;Wang, J.; Hu, X.; Ahmed, F.; Liu, Z.; Lu, Y.;Wang, L. UniTAB: Unifying Text and Box Outputs for Grounded Vision-Language Modeling. In Proceedings of the Computer Vision—ECCV, Tel Aviv, Israel, 23–27 October 2022.

• Zhao, C., Yang, J., & Xiong, W. et al. (2021). Two Generative Design Methods of Hospital Operating Department Layouts Based on Healthcare Systematic Layout Planning and Generative Adversarial Network. J. Shanghai Jiaotong Univ. (Sci.) 26, 103–115 https://doi.org/10.1007/s12204-021-2265-9.
• Zheng, H., & Yuan, P. F. (2021). A generative architectural and urban design method through artificial neural networks. Building and Environment, 205, 108178. https://doi.org/10.1016/j.buildenv.2021.108178.
• Zheng, H., An, K., Wei, J., & Ren, Y. (2020). Apartment Floor Plans Generation via Generative Adversarial Networks. In Proceedings of the RE: Anthropocene, Design in the Age of Humans: Proceedings of the 25th International Conference on Computer-Aided Architectural Design Research in Asia (CAADRIA 2020), Chulalongkorn University, Bangkok, Thailand. 10.52842/conf.caadria.2020.2.599.