Loic Landrieu

Unbalanced Optimal Transport enables accurate tree counting from noisy satellite data, outperforming traditional methods in dense forest environments.

UniverSat achieves sensor-agnostic feature extraction, outperforming traditional models in Earth Observation tasks by leveraging a flexible Universal Patch Encoder.

Unlock zero-shot geospatial reasoning by jointly embedding satellite imagery, street view, elevation, text, and lat/lon coordinates into a single space.

Attention's quadratic scaling problem? Solved: this new Polynomial Mixer (PoM) matches attention performance at linear complexity across diverse tasks.