Defenses in Adversarial Machine Learning: A Systematic Survey From the Lifecycle Perspective
IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026-01 · 48 (1) · 876-895
DOI: 10.1109/tpami.2025.3611340 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026-01. Verfasst von Baoyuan Wu und 9 weiteren Autor:innen. Die Arbeit wurde bisher 19-mal zitiert (Crossref) und nennt 266 Literaturangaben. Gefördert von: Basic and Applied Basic Research Foundation of Guangdong Province, Shenzhen Science and Technology Program, Sub-topic of Key R&D Projects of the Ministry of Science and Technology.
Abstract (Originalsprache)
Adversarial phenomena have been widely observed in machine learning (ML) systems, especially those using deep neural networks. These phenomena describe situations where ML systems may produce predictions that are inconsistent and incomprehensible to humans in certain specific cases. Such behavior poses a serious security threat to the practical application of ML systems. To exploit this vulnerability, several advanced attack paradigms have been developed, mainly including backdoor attacks, weight attacks, and adversarial examples. For each individual attack paradigm, various defense mechanisms have been proposed to enhance the robustness of models against the corresponding attacks. However, due to the independence and diversity of these defense paradigms, it is challenging to assess the overall robustness of an ML system against different attack paradigms. This survey aims to provide a systematic review of all existing defense paradigms from a unified lifecycle perspective. Specifically, we decompose a complete ML system into five stages: pre-training, training, post-training, deployment, and inference. We then present a clear taxonomy to categorize representative defense methods at each stage. The unified perspective and taxonomy not only help us analyze defense mechanisms but also enable us to understand the connections and differences among different defense paradigms. It inspires future research to develop more advanced and comprehensive defense strategies.
Förderung
- Basic and Applied Basic Research Foundation of Guangdong Province
- Shenzhen Science and Technology Program
- Sub-topic of Key R&D Projects of the Ministry of Science and Technology
- Longgang District Key Laboratory of Intelligent Digital Economy Security
- Guangdong Provincial Program
- National Natural Science Foundation of China
- CCF-Tencent Rhino-Bird Open Research Fund
Zitieren
APA
Wu, B., Zhu, M., Zheng, M., Zhu, Z., Wei, S., Zhang, M., Chen, H., Yuan, D., Liu, L., & Liu, Q. (2026). Defenses in Adversarial Machine Learning: A Systematic Survey From the Lifecycle Perspective. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(1), 876-895. https://doi.org/10.1109/tpami.2025.3611340
Vancouver
Wu B, Zhu M, Zheng M, Zhu Z, Wei S, Zhang M, et al. Defenses in Adversarial Machine Learning: A Systematic Survey From the Lifecycle Perspective. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026;48(1):876-895. doi:10.1109/tpami.2025.3611340
BibTeX
@article{Wu2026,
author = {Wu, Baoyuan and Zhu, Mingli and Zheng, Meixi and Zhu, Zihao and Wei, Shaokui and Zhang, Mingda and Chen, Hongrui and Yuan, Danni and Liu, Li and Liu, Qingshan},
title = {Defenses in Adversarial Machine Learning: A Systematic Survey From the Lifecycle Perspective},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
year = {2026},
volume = {48},
number = {1},
pages = {876--895},
doi = {10.1109/tpami.2025.3611340}
}
EUVIMED