UniverSat is a ViT-style backbone for Earth Observation built around a Universal Patch Encoder that handles arbitrary spatial, spectral, and temporal resolutions, and any sensor — with a single set of...

UNIGEOCLIP aligns text, aerial imagery, street-level views, DSMs, and geographic coordinates in a single unified embedding space via all-to-all contrastive learning.

This paper introduces the Polynomial Mixer (PoM), a novel token mixing mechanism with linear complexity that serves as a drop-in replacement for self-attention. PoM aggregates input tokens into a comp...

The growing availability of high-quality Earth Observation (EO) data enables accurate global land cover and crop type monitoring. However, the volume and heterogeneity of these datasets pose major pro...

Despite growing empirical evidence of bias amplification in machine learning, its theoretical foundations remain poorly understood. We develop a formal framework for majority-minority learning tasks, ...

Bi-temporal change detection at scale based on Very High Resolution (VHR) images is crucial for Earth monitoring. This remains poorly addressed so far: methods either require large volumes of annotate...