Analyzing the Adversarial Robustness and Interpretability of Deep SAR Classification Models: A Comprehensive Examination of Their Reliability

Spremljeno u:
Bibliografski detalji
Izdano u:Remote Sensing vol. 17, no. 11 (2025), p. 1943
Glavni autor: Chen Tianrui
Daljnji autori: Zhang Limeng, Guo Weiwei, Zhang Zenghui, Datcu Mihai
Izdano:
MDPI AG
Teme:
Online pristup:Citation/Abstract
Full Text + Graphics
Full Text - PDF
Oznake: Dodaj oznaku
Bez oznaka, Budi prvi tko označuje ovaj zapis!

MARC

LEADER 00000nab a2200000uu 4500
001 3217746045
003 UK-CbPIL
022 |a 2072-4292 
024 7 |a 10.3390/rs17111943  |2 doi 
035 |a 3217746045 
045 2 |b d20250101  |b d20251231 
084 |a 231556  |2 nlm 
100 1 |a Chen Tianrui  |u Shanghai Key Laboratory of Intelligent Sensing and Recognition, Shanghai Jiao Tong University, Shanghai 200240, China; chen_tianrui@sjtu.edu.cn (T.C.); zhanglimeng21@sjtu.edu.cn (L.Z.) 
245 1 |a Analyzing the Adversarial Robustness and Interpretability of Deep SAR Classification Models: A Comprehensive Examination of Their Reliability 
260 |b MDPI AG  |c 2025 
513 |a Journal Article 
520 3 |a Deep neural networks (DNNs) have shown strong performance in synthetic aperture radar (SAR) image classification. However, their “black-box” nature limits interpretability and poses challenges for robustness, which is critical for sensitive applications such as disaster assessment, environmental monitoring, and agricultural insurance. This study systematically evaluates the adversarial robustness of five representative DNNs (VGG11/16, ResNet18/101, and A-ConvNet) under a variety of attack and defense settings. Using eXplainable AI (XAI) techniques and attribution-based visualizations, we analyze how adversarial perturbations and adversarial training affect model behavior and decision logic. Our results reveal significant robustness differences across architectures, highlight interpretability limitations, and suggest practical guidelines for building more robust SAR classification systems. We also discuss challenges associated with large-scale, multi-class land use and land cover (LULC) classification under adversarial conditions. 
653 |a Environmental monitoring 
653 |a Accuracy 
653 |a Classification systems 
653 |a Datasets 
653 |a Deep learning 
653 |a Land use 
653 |a Network reliability 
653 |a Artificial neural networks 
653 |a Synthetic aperture radar 
653 |a Sensors 
653 |a Neural networks 
653 |a Classification 
653 |a Image classification 
653 |a Methods 
653 |a Radar imaging 
653 |a Land cover 
653 |a Explainable artificial intelligence 
653 |a Robustness 
700 1 |a Zhang Limeng  |u Shanghai Key Laboratory of Intelligent Sensing and Recognition, Shanghai Jiao Tong University, Shanghai 200240, China; chen_tianrui@sjtu.edu.cn (T.C.); zhanglimeng21@sjtu.edu.cn (L.Z.) 
700 1 |a Guo Weiwei  |u Center of Digital Innovation, Tongji University, Shanghai 200092, China; weiweiguo@tongji.edu.cn 
700 1 |a Zhang Zenghui  |u Shanghai Key Laboratory of Intelligent Sensing and Recognition, Shanghai Jiao Tong University, Shanghai 200240, China; chen_tianrui@sjtu.edu.cn (T.C.); zhanglimeng21@sjtu.edu.cn (L.Z.) 
700 1 |a Datcu Mihai  |u Research Center for Spatial Information (CEOSpaceTech), POLITEHNICA Bucharest, Bucharest 011061, Romania; mihai.datcu@upb.ro 
773 0 |t Remote Sensing  |g vol. 17, no. 11 (2025), p. 1943 
786 0 |d ProQuest  |t Advanced Technologies & Aerospace Database 
856 4 1 |3 Citation/Abstract  |u https://www.proquest.com/docview/3217746045/abstract/embedded/L8HZQI7Z43R0LA5T?source=fedsrch 
856 4 0 |3 Full Text + Graphics  |u https://www.proquest.com/docview/3217746045/fulltextwithgraphics/embedded/L8HZQI7Z43R0LA5T?source=fedsrch 
856 4 0 |3 Full Text - PDF  |u https://www.proquest.com/docview/3217746045/fulltextPDF/embedded/L8HZQI7Z43R0LA5T?source=fedsrch