Spectral–spatial mamba adversarial defense network for hyperspectral image classification

Deep learning models have obtained great success in hyperspectral image classification tasks. Nevertheless, they are usually vulnerable to adversarial attacks. Some existing works have been made to defend against adversarial attacks in HSI classification. These works primarily focus on lots of adver...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhongqiang Zhang, Ye Wang, Dahua Gao, Haoyong Li, Guangming Shi
Format: Article
Language:English
Published: Taylor & Francis Group 2025-08-01
Series:International Journal of Digital Earth
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/17538947.2025.2520480
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849224385005617152
author Zhongqiang Zhang
Ye Wang
Dahua Gao
Haoyong Li
Guangming Shi
author_facet Zhongqiang Zhang
Ye Wang
Dahua Gao
Haoyong Li
Guangming Shi
author_sort Zhongqiang Zhang
collection DOAJ
description Deep learning models have obtained great success in hyperspectral image classification tasks. Nevertheless, they are usually vulnerable to adversarial attacks. Some existing works have been made to defend against adversarial attacks in HSI classification. These works primarily focus on lots of adversarial samples and spatial relationships while overlooking the strong long-range dependencies from HSI. To alleviate this problem, we propose a novel spectral spatial mamba adversarial defense network (SSMADNet) for hyperspectral adversarial image classification. It includes a dense involution branch, a spectral mamba branch, and a spatial multiscale mamba branch. The dense involution branch extracts embedding features via three dense involution layers. The spectral mamba branch can learn the spectral sequence information from HSI adversarial samples. The spatial multiscale mamba branch can model the long-range interaction of the whole image. Finally, a spectral spatial feature enhancement module is designed to adaptively enhance useful spectral spatial features of HSI. Extensive experimental results demonstrate that on five HSI adversarial datasets, the proposed SSMADNet achieves higher classification accuracies than state-of-the-art adversarial defense methods. In particular, our method obtains best OA (93.80%) on the Botswana adversarial data, which is much higher than the suboptimal method (OA = 90.30%).
format Article
id doaj-art-9d3f02233a334e059bf5626305fbba9b
institution Kabale University
issn 1753-8947
1753-8955
language English
publishDate 2025-08-01
publisher Taylor & Francis Group
record_format Article
series International Journal of Digital Earth
spelling doaj-art-9d3f02233a334e059bf5626305fbba9b2025-08-25T11:25:07ZengTaylor & Francis GroupInternational Journal of Digital Earth1753-89471753-89552025-08-0118110.1080/17538947.2025.2520480Spectral–spatial mamba adversarial defense network for hyperspectral image classificationZhongqiang Zhang0Ye Wang1Dahua Gao2Haoyong Li3Guangming Shi4Peng Cheng Laboratory, Shenzhen, People’s Republic of ChinaPeng Cheng Laboratory, Shenzhen, People’s Republic of ChinaSchool of Artificial Intelligence, Xidian University, Xi’an, People’s Republic of ChinaSchool of Artificial Intelligence, Xidian University, Xi’an, People’s Republic of ChinaPeng Cheng Laboratory, Shenzhen, People’s Republic of ChinaDeep learning models have obtained great success in hyperspectral image classification tasks. Nevertheless, they are usually vulnerable to adversarial attacks. Some existing works have been made to defend against adversarial attacks in HSI classification. These works primarily focus on lots of adversarial samples and spatial relationships while overlooking the strong long-range dependencies from HSI. To alleviate this problem, we propose a novel spectral spatial mamba adversarial defense network (SSMADNet) for hyperspectral adversarial image classification. It includes a dense involution branch, a spectral mamba branch, and a spatial multiscale mamba branch. The dense involution branch extracts embedding features via three dense involution layers. The spectral mamba branch can learn the spectral sequence information from HSI adversarial samples. The spatial multiscale mamba branch can model the long-range interaction of the whole image. Finally, a spectral spatial feature enhancement module is designed to adaptively enhance useful spectral spatial features of HSI. Extensive experimental results demonstrate that on five HSI adversarial datasets, the proposed SSMADNet achieves higher classification accuracies than state-of-the-art adversarial defense methods. In particular, our method obtains best OA (93.80%) on the Botswana adversarial data, which is much higher than the suboptimal method (OA = 90.30%).https://www.tandfonline.com/doi/10.1080/17538947.2025.2520480Adversarial attacksspectral spatial mamba adversarial defense networkdense involution branchspatial multiscale mamba branch
spellingShingle Zhongqiang Zhang
Ye Wang
Dahua Gao
Haoyong Li
Guangming Shi
Spectral–spatial mamba adversarial defense network for hyperspectral image classification
International Journal of Digital Earth
Adversarial attacks
spectral spatial mamba adversarial defense network
dense involution branch
spatial multiscale mamba branch
title Spectral–spatial mamba adversarial defense network for hyperspectral image classification
title_full Spectral–spatial mamba adversarial defense network for hyperspectral image classification
title_fullStr Spectral–spatial mamba adversarial defense network for hyperspectral image classification
title_full_unstemmed Spectral–spatial mamba adversarial defense network for hyperspectral image classification
title_short Spectral–spatial mamba adversarial defense network for hyperspectral image classification
title_sort spectral spatial mamba adversarial defense network for hyperspectral image classification
topic Adversarial attacks
spectral spatial mamba adversarial defense network
dense involution branch
spatial multiscale mamba branch
url https://www.tandfonline.com/doi/10.1080/17538947.2025.2520480
work_keys_str_mv AT zhongqiangzhang spectralspatialmambaadversarialdefensenetworkforhyperspectralimageclassification
AT yewang spectralspatialmambaadversarialdefensenetworkforhyperspectralimageclassification
AT dahuagao spectralspatialmambaadversarialdefensenetworkforhyperspectralimageclassification
AT haoyongli spectralspatialmambaadversarialdefensenetworkforhyperspectralimageclassification
AT guangmingshi spectralspatialmambaadversarialdefensenetworkforhyperspectralimageclassification