Using a styled map from Mapbox Studio which only shows roads and road networks and the Mapbox static API, we collected the images of the road networks of more than 65K cities across the world.
Using the collected images of the city networks, we trained a Convolutional Autoencoder Neural Network, where similar to other autoencoders, the middle layer can be seen as a dense representation of the input data. Later these dense vectors can be used to map similar data points (here, cities) next to each other.
We then just simply use the learned vectors in a K-NN framework and find similar cities to the selected city by user.
Further visualizations and the collected data can be found at the project repo.
Developed by Vahid Moosavi at the Chair for Computer Aided Architectural Design (CAAD), ETH Zürich.