Abstract: While federated learning enables collaborative model training with preserved data locality, it remains vulnerable to evolving backdoor attacks that exploit its distributed architecture.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results