Tag: defenses against backdoor triggers in machine learning

Researchers at Donghua University have developed RFLDefense, a layer-wise defense framework that suppresses backdoor attacks in federated learning even when ...