Object Detection in Aerial Imagery

TL;DR

This paper evaluates various object detection models on aerial imagery, introduces modifications for improved performance, and compares their effectiveness considering challenges like high resolution and scale variation.

Contribution

It provides a comprehensive analysis of two-stage, one-stage, and attention-based detectors on aerial images, with specific modifications and a comparative study.

Findings

Weighted attention-based FPN improves detection accuracy.

Focal loss enhances one-stage detector performance.

Multi-scale attention benefits detection in aerial imagery.

Abstract

Object detection in natural images has achieved remarkable results over the years. However, a similar progress has not yet been observed in aerial object detection due to several challenges, such as high resolution images, instances scale variation, class imbalance etc. We show the performance of two-stage, one-stage and attention based object detectors on the iSAID dataset. Furthermore, we describe some modifications and analysis performed for different models - a) In two stage detector: introduced weighted attention based FPN, class balanced sampler and density prediction head. b) In one stage detector: used weighted focal loss and introduced FPN. c) In attention based detector: compare single,multi-scale attention and demonstrate effect of different backbones. Finally, we show a comparative study highlighting the pros and cons of different models in aerial imagery setting.