Explainable Analysis of Deep Learning Methods for Sar Image Classification

Shenghan Su, Ziteng Cui, Weiwei Guo, Zenghui Zhang, Wenxian Yu

2022IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium18 citationsDOI

Abstract

Deep learning methods exhibit outstanding performance in synthetic aperture radar (SAR) image interpretation tasks. However, these are black box models that limit the com-prehension of their predictions. Therefore, to meet this challenge, we have utilized explainable artificial intelli-gence (XAI) methods for the SAR image classification task. Specifically, we trained state-of-the-art convolutional neural networks for each polarization format on OpenSARUrban dataset and then investigate eight explanation methods to analyze the predictions of the CNN classifiers of SAR images. These XAI methods are also evaluated qualitatively and quantitatively which shows that Occlusion achieves the most reliable interpretation performance in terms of Max-Sensitivity but with a low-resolution explanation heatmap. The explanation results provide some insights into the in-ternal mechanism of black-box decisions for SAR image classification.

Topics & Concepts

Computer scienceArtificial intelligenceSynthetic aperture radarConvolutional neural networkDeep learningMargin (machine learning)Image (mathematics)Contextual image classificationPattern recognition (psychology)Black boxTask (project management)Machine learningInterpretation (philosophy)Programming languageEconomicsManagementAdversarial Robustness in Machine LearningAnomaly Detection Techniques and ApplicationsExplainable Artificial Intelligence (XAI)