Targeted universal adversarial examples for remote sensing
Researchers are focusing on the vulnerabilities of deep learning models for remote sensing; various attack methods have been proposed, including universal adversarial examples. Existing universal adversarial examples, however, are only designed to fool deep learning models rather than target specifi...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/165409 |
| _version_ | 1811686900514160640 |
|---|---|
| author | Bai, Tao Wang, Hao Wen, Bihan |
| author2 | School of Electrical and Electronic Engineering |
| author_facet | School of Electrical and Electronic Engineering Bai, Tao Wang, Hao Wen, Bihan |
| author_sort | Bai, Tao |
| collection | NTU |
| description | Researchers are focusing on the vulnerabilities of deep learning models for remote sensing; various attack methods have been proposed, including universal adversarial examples. Existing universal adversarial examples, however, are only designed to fool deep learning models rather than target specific goals, i.e., targeted attacks. To this end, we propose two variants of universal adversarial examples called targeted universal adversarial examples and source-targeted universal adversarial examples. Extensive experiments on three popular datasets showed strong attackability of the two targeted adversarial variants. We hope such strong attacks can inspire and motivate research on the defenses against adversarial examples in remote sensing. |
| first_indexed | 2024-10-01T05:07:46Z |
| format | Journal Article |
| id | ntu-10356/165409 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T05:07:46Z |
| publishDate | 2023 |
| record_format | dspace |
| spelling | ntu-10356/1654092023-03-31T16:02:34Z Targeted universal adversarial examples for remote sensing Bai, Tao Wang, Hao Wen, Bihan School of Electrical and Electronic Engineering Engineering::Electrical and electronic engineering Adversarial Examples Targeted Adversarial Attacks Researchers are focusing on the vulnerabilities of deep learning models for remote sensing; various attack methods have been proposed, including universal adversarial examples. Existing universal adversarial examples, however, are only designed to fool deep learning models rather than target specific goals, i.e., targeted attacks. To this end, we propose two variants of universal adversarial examples called targeted universal adversarial examples and source-targeted universal adversarial examples. Extensive experiments on three popular datasets showed strong attackability of the two targeted adversarial variants. We hope such strong attacks can inspire and motivate research on the defenses against adversarial examples in remote sensing. Ministry of Education (MOE) Published version This work was supported in part by the Ministry of Education, Republic of Singapore, under its Academic Research Fund Tier 1 Project RG61/22 and the Start-Up Grant. 2023-03-27T02:28:32Z 2023-03-27T02:28:32Z 2022 Journal Article Bai, T., Wang, H. & Wen, B. (2022). Targeted universal adversarial examples for remote sensing. Remote Sensing, 14(22), 5833-. https://dx.doi.org/10.3390/rs14225833 2072-4292 https://hdl.handle.net/10356/165409 10.3390/rs14225833 2-s2.0-85142718736 22 14 5833 en RG61/22 Remote Sensing © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). application/pdf |
| spellingShingle | Engineering::Electrical and electronic engineering Adversarial Examples Targeted Adversarial Attacks Bai, Tao Wang, Hao Wen, Bihan Targeted universal adversarial examples for remote sensing |
| title | Targeted universal adversarial examples for remote sensing |
| title_full | Targeted universal adversarial examples for remote sensing |
| title_fullStr | Targeted universal adversarial examples for remote sensing |
| title_full_unstemmed | Targeted universal adversarial examples for remote sensing |
| title_short | Targeted universal adversarial examples for remote sensing |
| title_sort | targeted universal adversarial examples for remote sensing |
| topic | Engineering::Electrical and electronic engineering Adversarial Examples Targeted Adversarial Attacks |
| url | https://hdl.handle.net/10356/165409 |
| work_keys_str_mv | AT baitao targeteduniversaladversarialexamplesforremotesensing AT wanghao targeteduniversaladversarialexamplesforremotesensing AT wenbihan targeteduniversaladversarialexamplesforremotesensing |