Project Gradient Descent Adversarial Attack against Multisource Remote Sensing Image Scene Classification
Authors:
Open Access
- 12 June 2021
- journal article
- research article
- Published by Wiley in Security and Communication Networks
- Vol. 2021, 1-13
Abstract
Deep learning technology (a deeper and optimized network structure) and remote sensing imaging (i.e., the more multisource and the more multicategory remote sensing data) have developed rapidly. Although the deep convolutional neural network (CNN) has achieved state-of-the-art performance on remote sensing image (RSI) scene classification, the existence of adversarial attacks poses a potential security threat to the RSI scene classification task based on CNN. The corresponding adversarial samples can be generated by adding a small perturbation to the original images. Feeding the CNN-based classifier with the adversarial samples leads to the classifier misclassify with high confidence. To achieve a higher attack success rate against scene classification based on CNN, we introduce the projected gradient descent method to generate adversarial remote sensing images. Then, we select several mainstream CNN-based classifiers as the attacked models to demonstrate the effectiveness of our method. The experimental results show that our proposed method can dramatically reduce the classification accuracy under untargeted and targeted attacks. Furthermore, we also evaluate the quality of the generated adversarial images by visual and quantitative comparisons. The results show that our method can generate the imperceptible adversarial samples and has a stronger attack ability for the RSI scene classification.Keywords
Funding Information
- Fundamental Research Funds for the Central Universities (3072020CFQ0602, 3072020CF0604, 3072020CFP0601, 3072021CF0609, KY1060020002, KY10600200008)
- 2019 Industrial Internet Innovation and Development Engineering (3072020CFQ0602, 3072020CF0604, 3072020CFP0601, 3072021CF0609, KY1060020002, KY10600200008)
This publication has 29 references indexed in Scilit:
- AID: A Benchmark Data Set for Performance Evaluation of Aerial Scene ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2017
- Deep Residual Learning for Image Recognition2016
- Effective and Efficient Midlevel Visual Elements-Oriented Land-Use Classification Using VHR Remote Sensing ImagesIEEE Transactions on Geoscience and Remote Sensing, 2015
- 2014 Index IEEE Transactions on Geoscience and Remote Sensing Vol. 52IEEE Transactions on Geoscience and Remote Sensing, 2014
- Bag of Lines (BoL) for Improved Aerial Scene RepresentationIEEE Geoscience and Remote Sensing Letters, 2014
- Pyramid of Spatial Relatons for Scene-Level Land Use ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2014
- Object detection in remote sensing imagery using a discriminatively trained mixture modelISPRS Journal of Photogrammetry and Remote Sensing, 2013
- Automatic landslide detection from remote-sensing imagery using a scene classification method based on BoVW and pLSAInternational Journal of Remote Sensing, 2012
- Evaluation of Morphological Texture Features for Mangrove Forest Mapping and Species Discrimination Using Multispectral IKONOS ImageryIEEE Geoscience and Remote Sensing Letters, 2009
- Distinctive Image Features from Scale-Invariant KeypointsInternational Journal of Computer Vision, 2004