GEO is a suite of open research projects building the data, models, and tools needed to measure, estimate, and improve educational access across low- and middle-income countries.
Projects
Each project is open-access and designed to be built upon by the research community.
GEO Dataset
A standardized, school-level administrative census spanning geocoordinates, personnel, infrastructure, and educational outcomes across low- and middle-income countries — built for GeoAI and spatial research.
TorchSDG4
Deep learning models for mapping SDG4 sub-indicators at high spatial resolution using satellite imagery — enabling education outcome estimation where EMIS data is sparse, delayed, or unreliable.
Model weights and data available upon publication.
GEDEx
Regional EMIS data collection and deep learning indicator estimation in collaboration with government and civil society partners in the Philippines, Cambodia, and Nigeria.
In active development with regional partners.
About GEO
GEO is a research program at the intersection of GeoAI, spatial data science, and international development — focused on making school-level data accessible, comparable, and useful across low- and middle-income countries.
All datasets are freely available for academic and humanitarian use. Models and code are released open-source on GitHub.