Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning
Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (19) · 21359-21367
DOI: 10.1609/aaai.v38i19.30131 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Tao Liu und 6 weiteren Autor:innen. Die Arbeit wurde bisher 43-mal zitiert (Crossref).
Abstract (Originalsprache)
Backdoors on federated learning will be diluted by subsequent benign updates. This is reflected in the significant reduction of attack success rate as iterations increase, ultimately failing. We use a new metric to quantify the degree of this weakened backdoor effect, called attack persistence. Given that research to improve this performance has not been widely noted, we propose a Full Combination Backdoor Attack (FCBA) method. It aggregates more combined trigger information for a more complete backdoor pattern in the global model. Trained backdoored global model is more resilient to benign updates, leading to a higher attack success rate on the test set. We test on three datasets and evaluate with two models across various settings. FCBA's persistence outperforms SOTA federated learning backdoor attacks. On GTSRB, post-attack 120 rounds, our attack success rate rose over 50% from baseline. The core code of our method is available at https://github.com/PhD-TaoLiu/FCBA.
Zitieren
APA
Liu, T., Zhang, Y., Feng, Z., Yang, Z., Xu, C., Man, D., & Yang, W. (2024). Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning. Proceedings of the AAAI Conference on Artificial Intelligence, 38(19), 21359-21367. https://doi.org/10.1609/aaai.v38i19.30131
Vancouver
Liu T, Zhang Y, Feng Z, Yang Z, Xu C, Man D, et al. Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(19):21359-21367. doi:10.1609/aaai.v38i19.30131
BibTeX
@article{Liu2024,
author = {Liu, Tao and Zhang, Yuhang and Feng, Zhu and Yang, Zhiqin and Xu, Chen and Man, Dapeng and Yang, Wu},
title = {Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2024},
volume = {38},
number = {19},
pages = {21359--21367},
doi = {10.1609/aaai.v38i19.30131}
}
EUVIMED