SaRNet: A Dataset for Deep Learning Assisted Search and Rescue with Satellite Imagery

TL;DR

SaRNet introduces a new satellite imagery dataset focused on small objects for Search and Rescue, along with baseline detection models and a specialized evaluation metric to advance deep learning applications in SaR operations.

Contribution

The paper presents a novel satellite imagery dataset for SaR, evaluates existing detection models on it, and proposes a new metric tailored for SaR scenarios.

Findings

Baseline models demonstrate potential for small object detection in SaR.

The new metric provides a more relevant evaluation for SaR applications.

Dataset facilitates future research in deep learning-assisted SaR.

Abstract

Access to high resolution satellite imagery has dramatically increased in recent years as several new constellations have entered service. High revisit frequencies as well as improved resolution has widened the use cases of satellite imagery to areas such as humanitarian relief and even Search and Rescue (SaR). We propose a novel remote sensing object detection dataset for deep learning assisted SaR. This dataset contains only small objects that have been identified as potential targets as part of a live SaR response. We evaluate the application of popular object detection models to this dataset as a baseline to inform further research. We also propose a novel object detection metric, specifically designed to be used in a deep learning assisted SaR setting.