Exploring Different Levels of Supervision for Detecting and Localizing Solar Panels on Remote Sensing Imagery

TL;DR

This paper compares different supervision levels for detecting and localizing solar panels in remote sensing images, highlighting the strengths and limitations of each approach in terms of accuracy and data requirements.

Contribution

It evaluates fully supervised, weakly supervised, and minimally supervised models for solar panel detection, providing insights into their performance and data efficiency.

Findings

Classifier achieves 0.79 F1-score in presence detection.

Object detector provides precise localization with 0.72 F1-score.

Fusion of models shows potential for improved accuracy.

Abstract

This study investigates object presence detection and localization in remote sensing imagery, focusing on solar panel recognition. We explore different levels of supervision, evaluating three models: a fully supervised object detector, a weakly supervised image classifier with CAM-based localization, and a minimally supervised anomaly detector. The classifier excels in binary presence detection (0.79 F1-score), while the object detector (0.72) offers precise localization. The anomaly detector requires more data for viable performance. Fusion of model results shows potential accuracy gains. CAM impacts localization modestly, with GradCAM, GradCAM++, and HiResCAM yielding superior results. Notably, the classifier remains robust with less data, in contrast to the object detector.