Towards Consistent Geospatial Representations: Aligning One Encoder to Many Foundation Models

The paper introduces Pangaea-Latents, an extension of the Pangaea benchmark that provides frozen embeddings from five geospatial foundation models (GFMs) across eleven datasets, showing that these embeddings are neither redundant nor disjoint. It then proposes GeoMixture, a framework that aligns a Vision Transformer (ViT) to multiple GFMs per dataset using a multi-representation alignment (mREPA) objective, resulting in an encoder that achieves the best average rank across nine Pangaea datasets while maintaining cross-task consistency.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

• rag
• benchmarks
• safety