MAPEX: Modality-Aware Pruning of Experts for Remote Sensing Foundation Models

TL;DR

MAPEX introduces a modality-aware pruning approach for remote sensing foundation models, enabling efficient, task-specific models that better match application modalities and improve performance across diverse remote sensing tasks.

Contribution

The paper presents MAPEX, a novel mixture-of-experts model with modality-conditioned routing and a pruning technique for tailored, efficient remote sensing models.

Findings

MAPEX outperforms existing models on various remote sensing datasets.

Modality-aware pruning simplifies fine-tuning and deployment.

The approach achieves strong results with reduced model complexity.

Abstract

Remote sensing data is commonly used for tasks such as flood mapping, wildfire detection, or land-use studies. For each task, scientists carefully choose appropriate modalities or leverage data from purpose-built instruments. Recent work on remote sensing foundation models pre-trains computer vision models on large amounts of remote sensing data. These large-scale models tend to focus on specific modalities, often optical RGB or multispectral data. For many important applications, this introduces a mismatch between the application modalities and the pre-training data. Moreover, the large size of foundation models makes them expensive and difficult to fine-tune on typically small datasets for each task. We address this mismatch with MAPEX, a remote sensing foundation model based on mixture-of-modality experts. MAPEX is pre-trained on multi-modal remote sensing data with a novel modality-conditioned token routing mechanism that elicits modality-specific experts. To apply the model on a specific task, we propose a modality aware pruning technique, which only retains experts specialized for the task modalities. This yields efficient modality-specific models while simplifying fine-tuning and deployment for the modalities of interest. We experimentally validate MAPEX on diverse remote sensing datasets and show strong performance compared to fully supervised training and state-of-the-art remote sensing foundation models. Code is available at https://github.com/HSG-AIML/MAPEX.