Masked Spatial-Spectral Autoencoders Are Excellent Hyperspectral Defenders

TL;DR

This paper introduces MSSA, a self-supervised masked autoencoder that enhances hyperspectral image analysis robustness against adversarial attacks by leveraging spectral and spatial features and graph-based pixel relationships.

Contribution

The paper proposes a novel masked spatial-spectral autoencoder with graph convolutional networks for improved adversarial defense in hyperspectral imaging.

Findings

MSSA outperforms state-of-the-art methods in robustness.

Effective in limited labeled data scenarios.

Demonstrates strong defense transferability.

Abstract

Deep learning methodology contributes a lot to the development of hyperspectral image (HSI) analysis community. However, it also makes HSI analysis systems vulnerable to adversarial attacks. To this end, we propose a masked spatial-spectral autoencoder (MSSA) in this paper under self-supervised learning theory, for enhancing the robustness of HSI analysis systems. First, a masked sequence attention learning module is conducted to promote the inherent robustness of HSI analysis systems along spectral channel. Then, we develop a graph convolutional network with learnable graph structure to establish global pixel-wise combinations.In this way, the attack effect would be dispersed by all the related pixels among each combination, and a better defense performance is achievable in spatial aspect.Finally, to improve the defense transferability and address the problem of limited labelled samples, MSSA employs spectra reconstruction as a pretext task and fits the datasets in a self-supervised manner.Comprehensive experiments over three benchmarks verify the effectiveness of MSSA in comparison with the state-of-the-art hyperspectral classification methods and representative adversarial defense strategies.