Active Ensemble Deep Learning for Polarimetric Synthetic Aperture Radar Image Classification

TL;DR

This paper introduces active ensemble deep learning (AEDL), leveraging snapshot disagreement in deep models to improve PolSAR image classification with limited labeled data, outperforming standard active learning methods.

Contribution

The novel AEDL method uses snapshot disagreement as an informativeness measure, enhancing classification accuracy with fewer training samples in PolSAR imagery.

Findings

AEDL outperforms standard active learning strategies.

Achieves same accuracy with 86% and 55% fewer samples.

Snapshots disagreement is a valuable indicator of data importance.

Abstract

Although deep learning has achieved great success in image classification tasks, its performance is subject to the quantity and quality of training samples. For classification of polarimetric synthetic aperture radar (PolSAR) images, it is nearly impossible to annotate the images from visual interpretation. Therefore, it is urgent for remote sensing scientists to develop new techniques for PolSAR image classification under the condition of very few training samples. In this letter, we take the advantage of active learning and propose active ensemble deep learning (AEDL) for PolSAR image classification. We first show that only 35\% of the predicted labels of a deep learning model's snapshots near its convergence were exactly the same. The disagreement between snapshots is non-negligible. From the perspective of multiview learning, the snapshots together serve as a good committee to evaluate the importance of unlabeled instances. Using the snapshots committee to give out the informativeness of unlabeled data, the proposed AEDL achieved better performance on two real PolSAR images compared with standard active learning strategies. It achieved the same classification accuracy with only 86% and 55% of the training samples compared with breaking ties active learning and random selection for the Flevoland dataset.