Improving Hyperspectral Adversarial Robustness Under Multiple Attacks

Nicholas Soucy, Salimeh Yasaei Sekeh

Keywords: Machine Learning, Semantic Segmentation, Ensemble Method, Neural Networks, Hyperspectral Data

TL;DR: In this work, the Adversarial Discriminator Ensemble Network (ADE-Net) is developed to improve adversarial detection and semantic segmentation robustness in the presence of multiple attacks on hyperspectral data under a unified model.

Abstract: Semantic segmentation models classifying hyperspectral images (HSI) are vulnerable to adversarial examples. Traditional approaches to adversarial robustness focus on training or retraining a single network on attacked data, however, in the presence of multiple attacks these approaches decrease in performance compared to networks trained individually on each attack. To combat this issue we propose an Adversarial Discriminator Ensemble Network (ADE-Net) which focuses on attack type detection and adversarial robustness under a unified model to preserve per data-type weight optimally while robustifiying the overall network. In the proposed method, a discriminator network is used to separate data by attack type into their specific attack-expert ensemble network.