Ensemble Learning techniques for object detection in high-resolution satellite images

TL;DR

This paper reviews ensembling techniques for object detection in very high resolution satellite images, highlighting their potential to improve detection performance in defense-related applications.

Contribution

It introduces the application of ensembling methods specifically for VHR satellite imagery in object detection, an area previously underexplored.

Findings

Ensembling improves detection accuracy in VHR satellite images.

Demonstrated ensembling benefits on vehicle detection in desert environments.

Provides a practical example of ensembling in operational remote sensing tasks.

Abstract

Ensembling is a method that aims to maximize the detection performance by fusing individual detectors. While rarely mentioned in deep-learning articles applied to remote sensing, ensembling methods have been widely used to achieve high scores in recent data science com-petitions, such as Kaggle. The few remote sensing articles mentioning ensembling mainly focus on mid resolution images and earth observation applications such as land use classification, but never on Very High Resolution (VHR) images for defense-related applications or object detection.This study aims at reviewing the most relevant ensembling techniques to be used for object detection on very high resolution imagery and shows an example of the value of such techniques on a relevant operational use-case (vehicle detection in desert areas).