Summary of Lurking in the Shadows: Unveiling Stealthy Backdoor Attacks Against Personalized Federated Learning, by Xiaoting Lyu et al.
Lurking in the shadows: Unveiling Stealthy Backdoor Attacks against Personalized Federated Learning
by Xiaoting Lyu, Yufei Han, Wei Wang, Jingkai Liu, Yongsheng Zhu, Guangquan Xu, Jiqiang Liu, Xiangliang Zhang
First submitted to arxiv on: 10 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed paper delves into the vulnerabilities of Personalized Federated Learning (PFL) to backdoor attacks. It demonstrates that while personalization can dilute poisoning effects and deploy server-end and client-end defense mechanisms, PFL fortified with these methods may offer a false sense of security. The authors introduce PFedBA, a stealthy attack strategy that optimizes trigger generation for seamless embedding into personalized local models. PFedBA achieves outstanding performance across 10 state-of-the-art PFL algorithms, defeating 6 existing defense mechanisms. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary PFL is a way for devices to work together and learn from each other without sharing their private data. However, this method has a problem when the data isn’t evenly distributed among the devices. To solve this issue, personalized FL (PFL) allows each device to create its own model tailored to its unique data. While PFL is promising, it’s still vulnerable to backdoor attacks. The authors of this study explored these vulnerabilities and found that even with defense mechanisms in place, PFL can be fooled by sophisticated attacks. |
Keywords
» Artificial intelligence » Embedding » Federated learning
