GeoMFD: Continual Drone-View Geo-Localization with Geometry-Aware Adapter and Margin-Field Distillation

2026-07-28

AI Summary

This work addresses the challenges of catastrophic forgetting and storage overhead in drone-view geolocalization when continuously learning from new environments. It formalizes, for the first time, the task of continual drone geolocalization and introduces GeoMFD, a method that integrates a cold-start bootstrapping strategy (CBS), a geometry-aware adapter (Geo-Adapter), and margin field distillation (MFD). Within a single-model continual updating framework, GeoMFD effectively preserves the geometric structural consistency of cross-view embeddings. Experimental results demonstrate that GeoMFD substantially mitigates catastrophic forgetting, achieving performance comparable to that of environment-specific ensemble models while simultaneously maintaining strong adaptability and memory retention.

Abstract

Existing drone-view geo-localization (DVGL) methods are mainly developed under a static training paradigm, where models are optimized for fixed environments with all training data available in advance. However, this paradigm is difficult to extend to real-world deployment, where drones may encounter diverse environments and require multiple environment-specific models, resulting in additional storage and model-selection costs. Directly adapting a single model to new environments also risks distorting previously learned cross-view embedding geometry and causing forgetting. To address these challenges, we formalize the continual drone-view geo-localization (C-DVGL) setting and propose GeoMFD, a geometry-aware continual adaptation method for DVGL. GeoMFD combines a cold-start bootstrapping strategy (CBS), a geometry-aware adapter (Geo-Adapter), and margin-field distillation (MFD) to balance adaptation and cross-view geometry preservation. CBS initializes a stable embedding space, Geo-Adapter enables environment adaptation through controlled residual corrections, and MFD preserves similarity margins between positive pairs and hard negatives to alleviate cross-view geometry forgetting. Extensive experiments demonstrate that GeoMFD effectively mitigates forgetting and achieves competitive performance with environment-specific DVGL methods using a single continuously updated model.

Problem

• continual learning
• drone-view geo-localization
• catastrophic forgetting
• cross-view geometry
• model adaptation

Innovation

• continual learning
• geo-localization
• geometry preservation
• adapter
• knowledge distillation