Abstract:We present a novel graph Transformer generative adversarial network (GTGAN) to learn effective graph node relations in an end-to-end fashion for challenging graph-constrained architectural layout generation tasks. The proposed graph-Transformer-based generator includes a novel graph Transformer encoder that combines graph convolutions and self-attentions in a Transformer to model both local and global interactions across connected and non-connected graph nodes. Specifically, the proposed connected node attention (CNA) and non-connected node attention (NNA) aim to capture the global relations across connected nodes and non-connected nodes in the input graph, respectively. The proposed graph modeling block (GMB) aims to exploit local vertex interactions based on a house layout topology. Moreover, we propose a new node classification-based discriminator to preserve the high-level semantic and discriminative node features for different house components. To maintain the relative spatial relationships between ground truth and predicted graphs, we also propose a novel graph-based cycle-consistency loss. Finally, we propose a novel self-guided pre-training method for graph representation learning. This approach involves simultaneous masking of nodes and edges at an elevated mask ratio (i.e., 40%) and their subsequent reconstruction using an asymmetric graph-centric autoencoder architecture. This method markedly improves the model's learning proficiency and expediency. Experiments on three challenging graph-constrained architectural layout generation tasks (i.e., house layout generation, house roof generation, and building layout generation) with three public datasets demonstrate the effectiveness of the proposed method in terms of objective quantitative scores and subjective visual realism. New state-of-the-art results are established by large margins on these three tasks.

Automated Layout Generation from Sites to Flats Using GAN and Transfer Learning

Automated building layout generation using deep learning and graph algorithms

House-GAN++: Generative Adversarial Layout Refinement Network towards Intelligent Computational Agent for Professional Architects

Graph Transformer GANs with Graph Masked Modeling for Architectural Layout Generation

Intelligent floor plan design of modular high-rise residential building based on graph-constrained generative adversarial networks

Automatic generation of architecture facade for historical urban renovation using generative adversarial network

ActFloor-GAN: Activity-Guided Adversarial Networks for Human-Centric Floorplan Design

Automated urban planning aware spatial hierarchies and human instructions

Automatic tower crane layout planning system for high-rise building construction using generative adversarial network

End‐to‐end generation of structural topology for complex architectural layouts with graph neural networks

Attribute-Conditioned Layout GAN for Automatic Graphic Design

Generative Deep Learning in Architectural Design

CRTGAN: Controllable Road Network Graphs Generation Via Transformer Based GAN

LayoutGAN: Generating Graphic Layouts with Wireframe Discriminators

LayoutGAN: Synthesizing Graphic Layouts with Vector-Wireframe Adversarial Networks

Automated Structural Design of Shear Wall Residential Buildings Using GAN-Based Machine Learning Algorithm

House-GAN: Relational Generative Adversarial Networks for Graph-constrained House Layout Generation

Intelligent layout generation based on deep generative models: A comprehensive survey

EL-GAN: Edge-Enhanced Generative Adversarial Network for Layout-to-Image Generation

LayoutDM: Transformer-based Diffusion Model for Layout Generation