Automated layout generation from sites to flats using GAN and transfer learning

• Published in: Automation in construction Vol. 166; p. 105668
• Main Authors: Wang, Lufeng, Zhou, Xuhong, Liu, Jiepeng, Cheng, Guozhong
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2024
• Subjects: AI-generated Content (AIGC); Floor plan; GAN; Layout generation; Transfer learning; Floor plan; AI-generated Content (AIGC); Layout generation; GAN; Transfer learning
• ISSN: 0926-5805
• DOI: 10.1016/j.autcon.2024.105668

Generating architectural layouts from sites to flats, encompassing site layouts (SLs), building layouts (BLs), and flat layouts (FLs), presents a complex process. Notably, the BL generation is challenging due to the small scale of data, making it difficult to train effective neural networks. This paper introduces an approach for generating layouts throughout the complete process. Initially, it proposes an enhanced generative adversarial network (GAN) combined with the transformer for Stable Diffusion (TranSD-GAN), considering design boundaries and requirements. Subsequently, for generating BLs with small-scale datasets, the paper proposes a stacking transfer learning method. Following this, image operations are conducted to support the flow of building information. Ultimately, BIM models are created at each stage. Through comparative experiments involving neural networks and generation cases, it is demonstrated that the proposed method significantly improves the generative capabilities of small-scale datasets and effectively aids designers throughout the layout design from sites to flats. •Architectural layouts from sites to flats are generated automatically.•An enhanced GAN with the transformer from Stable Diffusion is proposed.•Stacking transfer learning is proposed for a small-scale building layout dataset.•Image operations are performed to streamline building information flow.