FlowEO: Generative Unsupervised Domain Adaptation for Earth Observation

2025-12-01

AI Summary

Earth observation (EO) data exhibit significant domain shifts due to disparities in sensors, geographic regions, and acquisition times, severely limiting the cross-domain generalization of pretrained remote sensing models. To address this, we propose FlowEO—the first generative unsupervised domain adaptation framework tailored for remote sensing—introducing flow matching into EO image-to-image translation to achieve semantic-preserving, high-fidelity mapping from source to target domains. FlowEO supports challenging scenarios including SAR-optical cross-modal translation and post-disaster change detection, substantially improving generalization across downstream classification and semantic segmentation tasks. Extensive experiments on four benchmark EO datasets demonstrate that FlowEO outperforms existing image translation methods in domain adaptation effectiveness while maintaining superior or comparable perceptual quality.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Application Domains: Internet of Things, Sensor Networks & Smart Cities

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Security and Privacy: Data transparency and provenance
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

The increasing availability of Earth observation data offers unprecedented opportunities for large-scale environmental monitoring and analysis. However, these datasets are inherently heterogeneous, stemming from diverse sensors, geographical regions, acquisition times, and atmospheric conditions. Distribution shifts between training and deployment domains severely limit the generalization of pretrained remote sensing models, making unsupervised domain adaptation (UDA) crucial for real-world applications. We introduce FlowEO, a novel framework that leverages generative models for image-space UDA in Earth observation. We leverage flow matching to learn a semantically preserving mapping that transports from the source to the target image distribution. This allows us to tackle challenging domain adaptation configurations for classification and semantic segmentation of Earth observation images. We conduct extensive experiments across four datasets covering adaptation scenarios such as SAR to optical translation and temporal and semantic shifts caused by natural disasters. Experimental results demonstrate that FlowEO outperforms existing image translation approaches for domain adaptation while achieving on-par or better perceptual image quality, highlighting the potential of flow-matching-based UDA for remote sensing.

Problem

• Addresses domain shifts in Earth observation data for model generalization
• Uses generative models to adapt images between source and target domains
• Improves classification and segmentation across sensors, times, and disasters

Innovation

• FlowEO uses flow matching for domain adaptation
• It translates images from source to target distributions
• Framework improves classification and segmentation accuracy