Scene-to-Patch Earth Observation: Multiple Instance Learning for Land Cover Classification

TL;DR

This paper introduces Scene-to-Patch models using Multiple Instance Learning for land cover classification, enabling effective predictions with only scene-level labels, thus reducing dataset annotation effort and expanding application potential.

Contribution

The study presents a novel MIL-based approach for land cover classification that requires only high-level scene labels, facilitating faster dataset creation and broader use.

Findings

Outperforms non-MIL baselines on DeepGlobe-LCC dataset

Achieves accurate scene- and patch-level predictions

Reduces need for detailed annotations

Abstract

Land cover classification (LCC), and monitoring how land use changes over time, is an important process in climate change mitigation and adaptation. Existing approaches that use machine learning with Earth observation data for LCC rely on fully-annotated and segmented datasets. Creating these datasets requires a large amount of effort, and a lack of suitable datasets has become an obstacle in scaling the use of LCC. In this study, we propose Scene-to-Patch models: an alternative LCC approach utilising Multiple Instance Learning (MIL) that requires only high-level scene labels. This enables much faster development of new datasets whilst still providing segmentation through patch-level predictions, ultimately increasing the accessibility of using LCC for different scenarios. On the DeepGlobe-LCC dataset, our approach outperforms non-MIL baselines on both scene- and patch-level prediction. This work provides the foundation for expanding the use of LCC in climate change mitigation methods for technology, government, and academia.