Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar Images
Synthetic aperture radar (SAR) automatic target recognition (ATR) systems based on deep neural network models are vulnerable to adversarial examples. Existing SAR adversarial attack algorithms require access to the network structure, parameters, and training data, which are often inaccessible in rea...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10768976/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841533400787189760 |
|---|---|
| author | Xuanshen Wan Wei Liu Chaoyang Niu Wanjie Lu Yuanli Li |
| author_facet | Xuanshen Wan Wei Liu Chaoyang Niu Wanjie Lu Yuanli Li |
| author_sort | Xuanshen Wan |
| collection | DOAJ |
| description | Synthetic aperture radar (SAR) automatic target recognition (ATR) systems based on deep neural network models are vulnerable to adversarial examples. Existing SAR adversarial attack algorithms require access to the network structure, parameters, and training data, which are often inaccessible in real-world scenarios. To address this problem, this study proposes an intermediate layer feature attack algorithm that does not rely on training data for the adversary model. Electromagnetic simulation is used to obtain the simulated SAR local data domain during the training stage. A lightweight generator, TinyResNet, is introduced to quickly construct adversarial examples through a one-step mapping process. In addition, the transferability of these examples across different models is improved by eliminating the intermediate layer features of the model. Finally, a domain-agnostic feature attention module is utilized to reduce discrepancies between different data domains from a model perspective, further improving the transferability of adversarial examples across domains. Experimental results on five SAR datasets of ground vehicles, ships, and scene types demonstrate that the proposed algorithm outperforms 13 mainstream adversarial attack algorithms in terms of cross-model and cross-data domain transferability. In particular, the proposed method improves the cross-domain attack success rate by 43.74%–48.88% on the MSTAR dataset. |
| format | Article |
| id | doaj-art-4460a0a33a4c4a9faf5e9c1eb95a0c7d |
| institution | Kabale University |
| issn | 1939-1404 2151-1535 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj-art-4460a0a33a4c4a9faf5e9c1eb95a0c7d2025-01-16T00:00:25ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01181638165510.1109/JSTARS.2024.350737410768976Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar ImagesXuanshen Wan0https://orcid.org/0009-0005-4146-3567Wei Liu1https://orcid.org/0000-0002-3395-6696Chaoyang Niu2https://orcid.org/0000-0001-5144-8815Wanjie Lu3https://orcid.org/0000-0002-7210-7670Yuanli Li4https://orcid.org/0009-0004-3272-9952Information Engineering University, Zhengzhou, ChinaInformation Engineering University, Zhengzhou, ChinaInformation Engineering University, Zhengzhou, ChinaInformation Engineering University, Zhengzhou, ChinaInformation Engineering University, Zhengzhou, ChinaSynthetic aperture radar (SAR) automatic target recognition (ATR) systems based on deep neural network models are vulnerable to adversarial examples. Existing SAR adversarial attack algorithms require access to the network structure, parameters, and training data, which are often inaccessible in real-world scenarios. To address this problem, this study proposes an intermediate layer feature attack algorithm that does not rely on training data for the adversary model. Electromagnetic simulation is used to obtain the simulated SAR local data domain during the training stage. A lightweight generator, TinyResNet, is introduced to quickly construct adversarial examples through a one-step mapping process. In addition, the transferability of these examples across different models is improved by eliminating the intermediate layer features of the model. Finally, a domain-agnostic feature attention module is utilized to reduce discrepancies between different data domains from a model perspective, further improving the transferability of adversarial examples across domains. Experimental results on five SAR datasets of ground vehicles, ships, and scene types demonstrate that the proposed algorithm outperforms 13 mainstream adversarial attack algorithms in terms of cross-model and cross-data domain transferability. In particular, the proposed method improves the cross-domain attack success rate by 43.74%–48.88% on the MSTAR dataset.https://ieeexplore.ieee.org/document/10768976/Adversarial exampleautomatic target recognition (ATR)deep neural network (DNN)synthetic aperture radar (SAR)transferability |
| spellingShingle | Xuanshen Wan Wei Liu Chaoyang Niu Wanjie Lu Yuanli Li Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar Images IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Adversarial example automatic target recognition (ATR) deep neural network (DNN) synthetic aperture radar (SAR) transferability |
| title | Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar Images |
| title_full | Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar Images |
| title_fullStr | Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar Images |
| title_full_unstemmed | Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar Images |
| title_short | Enhancing Adversarial Transferability With Intermediate Layer Feature Attack on Synthetic Aperture Radar Images |
| title_sort | enhancing adversarial transferability with intermediate layer feature attack on synthetic aperture radar images |
| topic | Adversarial example automatic target recognition (ATR) deep neural network (DNN) synthetic aperture radar (SAR) transferability |
| url | https://ieeexplore.ieee.org/document/10768976/ |
| work_keys_str_mv | AT xuanshenwan enhancingadversarialtransferabilitywithintermediatelayerfeatureattackonsyntheticapertureradarimages AT weiliu enhancingadversarialtransferabilitywithintermediatelayerfeatureattackonsyntheticapertureradarimages AT chaoyangniu enhancingadversarialtransferabilitywithintermediatelayerfeatureattackonsyntheticapertureradarimages AT wanjielu enhancingadversarialtransferabilitywithintermediatelayerfeatureattackonsyntheticapertureradarimages AT yuanlili enhancingadversarialtransferabilitywithintermediatelayerfeatureattackonsyntheticapertureradarimages |