IPBA: Imperceptible Perturbation Backdoor Attack in Federated Self-Supervised Learning
This addresses security risks in federated self-supervised learning, an incremental improvement for stealthier backdoor attacks in decentralized AI systems.
The paper tackles the vulnerability of federated self-supervised learning to backdoor attacks by proposing IPBA, an imperceptible perturbation method that decouples feature distributions and uses Sliced-Wasserstein distance, achieving significant performance gains and robustness over existing methods in experiments.
Federated self-supervised learning (FSSL) combines the advantages of decentralized modeling and unlabeled representation learning, serving as a cutting-edge paradigm with strong potential for scalability and privacy preservation. Although FSSL has garnered increasing attention, research indicates that it remains vulnerable to backdoor attacks. Existing methods generally rely on visually obvious triggers, which makes it difficult to meet the requirements for stealth and practicality in real-world deployment. In this paper, we propose an imperceptible and effective backdoor attack method against FSSL, called IPBA. Our empirical study reveals that existing imperceptible triggers face a series of challenges in FSSL, particularly limited transferability, feature entanglement with augmented samples, and out-of-distribution properties. These issues collectively undermine the effectiveness and stealthiness of traditional backdoor attacks in FSSL. To overcome these challenges, IPBA decouples the feature distributions of backdoor and augmented samples, and introduces Sliced-Wasserstein distance to mitigate the out-of-distribution properties of backdoor samples, thereby optimizing the trigger generation process. Our experimental results on several FSSL scenarios and datasets show that IPBA significantly outperforms existing backdoor attack methods in performance and exhibits strong robustness under various defense mechanisms.