City Foundation Models for Learning General Purpose Representations from OpenStreetMap

TL;DR

This paper introduces CityFM, a self-supervised foundation model trained on open geospatial data from OpenStreetMap, capable of generating multimodal representations for urban entities to improve various city-related tasks.

Contribution

CityFM is the first to leverage open geospatial data for training a multimodal foundation model tailored for urban applications, addressing data heterogeneity and enhancing downstream task performance.

Findings

CityFM outperforms or matches specialized algorithms in multiple tasks.

The model effectively integrates spatial, visual, and textual data.

Qualitative analysis shows meaningful entity representations.

Abstract

Pre-trained Foundation Models (PFMs) have ushered in a paradigm-shift in Artificial Intelligence, due to their ability to learn general-purpose representations that can be readily employed in a wide range of downstream tasks. While PFMs have been successfully adopted in various fields such as Natural Language Processing and Computer Vision, their capacity in handling geospatial data and answering urban questions remains limited. This can be attributed to the intrinsic heterogeneity of geospatial data, which encompasses different data types, including points, segments and regions, as well as multiple information modalities, such as a spatial position, visual characteristics and textual annotations. The proliferation of Volunteered Geographic Information initiatives, and the ever-increasing availability of open geospatial data sources, like OpenStreetMap, which is freely accessible globally, unveil a promising opportunity to bridge this gap. In this paper, we present CityFM, a self-supervised framework to train a foundation model within a selected geographical area of interest, such as a city. CityFM relies solely on open data from OSM, and produces multimodal representations of entities of different types, incorporating spatial, visual, and textual information. We analyse the entity representations generated using our foundation models from a qualitative perspective, and conduct quantitative experiments on road, building, and region-level downstream tasks. We compare its results to algorithms tailored specifically for the respective applications. In all the experiments, CityFM achieves performance superior to, or on par with, the baselines.