Adversarially Robust Hyperspectral Image Classification via Random Spectral Sampling and Spectral Shape Encoding

被引:10
|
作者
Park, Sungjune [1 ]
Lee, Hong Joo [1 ]
Ro, Yong Man [1 ]
机构
[1] Korea Adv Inst Sci & Technol KAIST, Sch Elect Engn, Daejeon 34141, South Korea
关键词
Training; Robustness; Spectral shape; Shape; Encoding; Feature extraction; Hyperspectral imaging; Adversarial robustness; hyperspectral image classification; random spectral sampling; spectral shape encoding; NEURAL-NETWORKS; ATTACKS; THREAT;
D O I
10.1109/ACCESS.2021.3076225
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Although the hyperspectral image (HSI) classification has adopted deep neural networks (DNNs) and shown remarkable performances, there is a lack of studies of the adversarial vulnerability for the HSI classifications. In this paper, we propose a novel HSI classification framework robust to adversarial attacks. To this end, we focus on the unique spectral characteristic of HSIs (i.e., distinctive spectral patterns of materials). With the spectral characteristic, we present the random spectral sampling and spectral shape feature encoding for the robust HSI classification. For the random spectral sampling, spectral bands are randomly sampled from the entire spectrum for each pixel of the input HSI. Also, the overall spectral shape information, which is robust to adversarial attacks, is fed into the shape feature extractor to acquire the spectral shape feature. Then, the proposed framework can provide the adversarial robustness of HSI classifiers via randomization effects and spectral shape feature encoding. To the best of our knowledge, the proposed framework is the first work dealing with the adversarial robustness in the HSI classification. In experiments, we verify that our framework improves the adversarial robustness considerably under diverse adversarial attack scenarios, and outperforms the existing adversarial defense methods.
引用
收藏
页码:66791 / 66804
页数:14
相关论文
共 50 条
  • [41] A fuzzy spectral clustering algorithm for hyperspectral image classification
    Li, Kang
    Xu, Jindong
    Zhao, Tianyu
    Liu, Zhaowei
    IET IMAGE PROCESSING, 2021, 15 (12) : 2810 - 2817
  • [42] Sparse Representations for the Spectral–Spatial Classification of Hyperspectral Image
    Mohamed Ali Hamdi
    Rafika Ben Salem
    Journal of the Indian Society of Remote Sensing, 2019, 47 : 923 - 929
  • [43] Spectral Swin Transformer Network for Hyperspectral Image Classification
    Liu, Baisen
    Liu, Yuanjia
    Zhang, Wulin
    Tian, Yiran
    Kong, Weili
    REMOTE SENSING, 2023, 15 (15)
  • [44] Spatial-Spectral Transformer for Hyperspectral Image Classification
    He, Xin
    Chen, Yushi
    Lin, Zhouhan
    REMOTE SENSING, 2021, 13 (03) : 1 - 22
  • [45] Spectral-Spatial Constraint Hyperspectral Image Classification
    Ji, Rongrong
    Gao, Yue
    Hong, Richang
    Liu, Qiong
    Tao, Dacheng
    Li, Xuelong
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2014, 52 (03): : 1811 - 1824
  • [46] Subpixel Spectral Variability Network for Hyperspectral Image Classification
    Han, Zhu
    Yang, Jin
    Gao, Lianru
    Zeng, Zhiqiang
    Zhang, Bing
    Chanussot, Jocelyn
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2025, 63
  • [47] Spectral-Spatial Response for Hyperspectral Image Classification
    Wei, Yantao
    Zhou, Yicong
    Li, Hong
    REMOTE SENSING, 2017, 9 (03):
  • [48] SPECTRAL PARTITIONING FOR HYPERSPECTRAL REMOTE SENSING IMAGE CLASSIFICATION
    Liu, Yi
    Li, Jun
    Plaza, Antonio
    Bioucas-Dias, Jose
    Cuartero, Aurora
    Garcia Rodriguez, Pablo
    2014 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS), 2014, : 3434 - 3437
  • [49] HYPERSPECTRAL IMAGE CLASSIFICATION BASED ON SPECTRAL AND GEOMETRICAL FEATURES
    Luo, Bin
    Chanussot, Jocelyn
    2009 IEEE INTERNATIONAL WORKSHOP ON MACHINE LEARNING FOR SIGNAL PROCESSING, 2009, : 246 - 251
  • [50] Spatial-Spectral BERT for Hyperspectral Image Classification
    Ashraf, Mahmood
    Zhou, Xichuan
    Vivone, Gemine
    Chen, Lihui
    Chen, Rong
    Majdard, Reza Seifi
    REMOTE SENSING, 2024, 16 (03)