Skip to yearly menu bar Skip to main content

Workshop: Workshop on Distribution Shifts: New Frontiers with Foundation Models

Exploring Generalisability of Self-Distillation with No Labels for SAR-Based Vegetation Prediction

Laura Martínez-Ferrer · Anna Jungbluth · Joseph Alejandro Gallego Mejia · Matthew Allen · Francisco Dorr · Freddie Kalaitzis · Raul Ramos-Pollán

Keywords: [ embeddings ] [ SAR ] [ Vegetation percentage ] [ DINO ] [ SSL ]


In this work we pre-train a DINO-ViT based model using two Synthetic Aperture Radar datasets (S1GRD or GSSIC) across three regions (China, Conus, Europe). We fine-tune the models on smaller labeled datasets to predict vegetation percentage, and empirically study the connection between the embedding space of the models and their ability to generalize across diverse geographic regions and to unseen data. For S1GRD, embedding spaces of different regions are clearly separated, while GSSIC's overlaps. Positional patterns remain during fine-tuning, and greater distances in embeddings often result in higher errors for unfamiliar regions. With this, our work increases our understanding of generalizability for self-supervised models applied to remote sensing.

Chat is not available.