Unsupervised Domain Adaptation for Enhanced Radiometer Image Precipitation Estimation using Conditional Flow Matching

2026-10-01

AI Summary

This study addresses the limited accuracy of precipitation estimation from satellite radiometer imagery caused by scarce labeled data and domain shift. To overcome this, it proposes an unsupervised domain adaptation framework based on conditional flow matching, which leverages deterministic ordinary differential equations to construct bijective transformations. This approach enables precise alignment of data across different satellite instruments during cross-domain adaptation while fully preserving essential physical information. By integrating deep generative networks, the method achieves distribution transfer without requiring target domain annotations. Experiments on the GPM-Core constellation demonstrate that the proposed framework significantly improves precipitation estimation accuracy from radiometer images, offering an efficient unsupervised solution for cross-sensor remote sensing tasks.

Abstract

Deep generative networks have recently achieved unprecedented performance in precise image and video editing using sophisticated textual prompts. However, the effectiveness of such models heavily depends on access to very large supervised and annotated image datasets, which can be very difficult to obtain. This is particularly true for satellite instruments, which very rarely overlap with labelled data, and suffer from domain shifts in the rare occasions they do. In this paper, we investigate the potential of flow matching models for unsupervised domain adaptation of satellite radiometer images. Our main contribution is a novel unsupervised method that achieves precise domain alignment by leveraging parts of the deterministic ordinary differential equations in flow matching models, conditioned on different satellite instruments. A key strength of our approach is its ability to preserve essential information while adapting across any domains since the perturbations are in theory bijective. Extensive experiments conducted on the GPM-Core constellation show the benefit of our conditional domain adaptation, particularly in improving rain precipitation estimation from radiometer imagery.

Problem

• Unsupervised Domain Adaptation
• Satellite Radiometer Images
• Precipitation Estimation
• Domain Shift

Innovation

• Unsupervised Domain Adaptation
• Conditional Flow Matching
• Radiometer Image
• Precipitation Estimation
• Ordinary Differential Equations