Modeling urbanization patterns with generative adversarial networks
This addresses the problem of modeling urbanization patterns for urban planners and researchers, but it appears incremental as it builds on existing GAN methods with new data.
The study tackled simulating realistic urban patterns by proposing a new method using Generative Adversarial Networks trained on global urban land-use data, resulting in a synthetic urban universe that qualitatively reproduces complex spatial organization and quantitatively recovers key high-level urban spatial metrics.
In this study we propose a new method to simulate hyper-realistic urban patterns using Generative Adversarial Networks trained with a global urban land-use inventory. We generated a synthetic urban "universe" that qualitatively reproduces the complex spatial organization observed in global urban patterns, while being able to quantitatively recover certain key high-level urban spatial metrics.