Rapid Adaptation of Earth Observation Foundation Models for Segmentation

TL;DR

This paper demonstrates that Low-Rank Adaptation (LoRA) effectively fine-tunes Earth Observation models for flood segmentation, achieving high accuracy with reduced computational costs and improved out-of-distribution generalization.

Contribution

It introduces the application of LoRA for efficient, high-performance flood segmentation in Earth Observation models, outperforming full fine-tuning methods.

Findings

LoRA improves F1 score by 6.66 points over baseline

LoRA reduces computational costs significantly

LoRA enhances out-of-distribution generalization

Abstract

This study investigates the efficacy of Low-Rank Adaptation (LoRA) in fine-tuning Earth Observation (EO) foundation models for flood segmentation. We hypothesize that LoRA, a parameter-efficient technique, can significantly accelerate the adaptation of large-scale EO models to this critical task while maintaining high performance. We apply LoRA to fine-tune a state-of-the-art EO foundation model pre-trained on diverse satellite imagery, using a curated dataset of flood events. Our results demonstrate that LoRA-based fine-tuning (r-256) improves F1 score by 6.66 points and IoU by 0.11 compared to a frozen encoder baseline, while significantly reducing computational costs. Notably, LoRA outperforms full fine-tuning, which proves computationally infeasible on our hardware. We further assess generalization through out-of-distribution (OOD) testing on a geographically distinct flood event. While LoRA configurations show improved OOD performance over the baseline. This work contributes to research on efficient adaptation of foundation models for specialized EO tasks, with implications for rapid response systems in disaster management. Our findings demonstrate LoRA's potential for enabling faster deployment of accurate flood segmentation models in resource-constrained, time-critical scenarios.