Understanding human-environment interactions and monitoring infrastructure changes are central challenges across several domains, including environmental science, urban studies, and conflict monitoring. At ECEO, we develop machine learning-based algorithms to track changes visible through remote sensing observations (e.g. VHR optical sensors, Nighttime Lights imagery). We are currently working on two distinct projects:
- The detection of infrastructure changes in urban areas, in collaboration with AXA.
- In collaboration with the Red Cross, we develop AI-driven tools to support crisis analysis in rural areas.
To ensure that our methods remain effective in a constantly changing world, we also study adaptive frameworks that allow models to detect previously unseen changes, leveraging the large volume of high-resolution, multi-temporal images made available by modern remote sensing technologies.
Papers
- BĂ©chaz, M., Dalsasso, E., TomoiagÄ, C., Detyniecki, M., and Tuia, D. Self-supervised Change Detection via Cooperative Learning: A Two-Player Model. In 2025 Joint Urban Remote Sensing Event (JURSE) (pp. 1-4). IEEE 2025 (paper).
- Metzger, N., Vargas-Muñoz, J.E., Daudt, R.C., Kellenberger, B., Whelan, T.T.T., Ofli, F., Imran, M., Schindler, K., and Tuia, D. Fine-grained population mapping from coarse census counts and open geodata. Scientific Reports, 12(1), p.20085, 2022 (paper).
