Article: Intelligent optimisation of traditional village element layout using generative adversarial networks Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.30 pp.60 - 80 Abstract: The conservation and optimisation of traditional villages has progressively been a major focus for intelligent applications with the fast development of artificial intelligence and information technology. Combining spatial multi-dimensional constraints to intelligibly optimise the layout of traditional villages through the adversarial training of generators and discriminators, this paper proposes a traditional village element layout optimisation method GAN-PLS based on generative adversarial network (GAN) model. Introduced into the GAN-PLS model, the gradient penalty technique helps to increase the stability of the training process and optimisation effect. Particularly outperforming conventional optimisation techniques in terms of convergence speed, generation effect and stability, the GAN-PLS model shows good performance in spatial layout optimisation, cultural element retention and ecological conservation by means of comparison studies. At last, this work addresses the shortcomings of the model and suggests future directions like dataset expansion, computational efficiency enhancement, and multi-dimensional constraints addition. Inderscience Publishers - linking academia, business and industry through research

Title: Intelligent optimisation of traditional village element layout using generative adversarial networks

Authors: Bin Yu; Ying Yang; Xie Liu

Addresses: College of Landscape Architecture, Central South University of Forestry and Technology, Changsha 410004, China; College of Tourism and Urban-Rural Planning, Xichang University, Xichang 615013, China ' China Construction Fifth Engineering Division Corporation Limited, Changsha 410001, China ' College of Tourism and Urban-Rural Planning, Xichang University, Xichang 615013, China; Faculty of Design and Architecture, University Putra Malaysia, Selangor 43400, Malaysia

Abstract: The conservation and optimisation of traditional villages has progressively been a major focus for intelligent applications with the fast development of artificial intelligence and information technology. Combining spatial multi-dimensional constraints to intelligibly optimise the layout of traditional villages through the adversarial training of generators and discriminators, this paper proposes a traditional village element layout optimisation method GAN-PLS based on generative adversarial network (GAN) model. Introduced into the GAN-PLS model, the gradient penalty technique helps to increase the stability of the training process and optimisation effect. Particularly outperforming conventional optimisation techniques in terms of convergence speed, generation effect and stability, the GAN-PLS model shows good performance in spatial layout optimisation, cultural element retention and ecological conservation by means of comparison studies. At last, this work addresses the shortcomings of the model and suggests future directions like dataset expansion, computational efficiency enhancement, and multi-dimensional constraints addition.

Keywords: generative adversarial network; GAN; intelligent optimisation of layout; traditional village; gradient penalty.

DOI: 10.1504/IJICT.2025.147765

International Journal of Information and Communication Technology, 2025 Vol.26 No.30, pp.60 - 80

Received: 02 Jun 2025
Accepted: 16 Jun 2025
Published online: 30 Jul 2025 *

Title: Intelligent optimisation of traditional village element layout using generative adversarial networks

Keep up-to-date