CartoVec: General-Purpose Location Embeddings from Open Map Data
Abstract
Geographic coordinates are common inputs in applied machine learning, but latitude and longitude describe position rather than the surrounding environment. Image-grounded location encoders learn this context from satellite or ground-level imagery; structured open map data offers another source of supervision whose utility for continuous coordinate encoders remains less explored. We introduce CartoVec, a coordinate encoder trained to predict Overture-derived land cover, land use, transportation, place and connectivity targets at multiple spatial scales, without image inputs or downstream-task labels. We evaluate frozen embeddings on Census outcomes, Airbnb prices and Philadelphia crime in the US Northeast using spatial block cross-validation. CartoVec has the highest mean performance on most Census outcomes and the highest average across the full suite, with several comparisons close across folds. The findings support open map data as a practical source of reusable location representations with a coordinate-to-vector interface that requires no map lookup at inference. The open-weight model is available at https://huggingface.co/gursoyfurkan/CartoVec.
Related
- Purpose-driven polymers: Mechanistic insights and empirical advances in inverse vulcanized polysulfides for mercury remediation
- Understanding the Global Report on Collective Progress in the Implementation of the Kunming-Montreal Global Biodiversity Framework: Purpose, Content, and Science-Policy Engagement
- Bridging forest inventory gaps in remote and inaccessible temperate mountain forests using AlphaEarth embeddings and model-based inference
- Methodological Bias between Trajectory Geometric Co-location and Air-Parcel Lagrangian Co-location: Evidence from UAV Re-penetration Observations over the Western Sichuan Plateau
- Fit for purpose evapotranspiration measurements: A multi-site comparison of lower-cost approaches across croplands and shrublands
BibTeX
@misc{gursoy2026cartovec,
title = {{CartoVec: General-Purpose Location Embeddings from Open Map Data}},
author = {Furkan Gursoy},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.17190},
eprint = {2610.17190},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.17190},
}
RIS
TY - UNPB
TI - CartoVec: General-Purpose Location Embeddings from Open Map Data
AU - Furkan Gursoy
PY - 2026
DA - 2026/10/12
PB - arXiv
DO - 10.48550/arxiv.2610.17190
UR - https://arxiv.org/abs/2610.17190
AB - Geographic coordinates are common inputs in applied machine learning, but latitude and longitude describe position rather than the surrounding environment. Image-grounded location encoders learn this context from satellite or ground-level imagery; structured open map data offers another source of supervision whose utility for continuous coordinate encoders remains less explored. We introduce CartoVec, a coordinate encoder trained to predict Overture-derived land cover, land use, transportation, place and connectivity targets at multiple spatial scales, without image inputs or downstream-task labels. We evaluate frozen embeddings on Census outcomes, Airbnb prices and Philadelphia crime in the US Northeast using spatial block cross-validation. CartoVec has the highest mean performance on most Census outcomes and the highest average across the full suite, with several comparisons close across folds. The findings support open map data as a practical source of reusable location representations with a coordinate-to-vector interface that requires no map lookup at inference. The open-weight model is available at https://huggingface.co/gursoyfurkan/CartoVec.
ER -