Dual-Masking Framework against Two-Sided Model Attacks in Federated Learning | IEEE Conference Publication | IEEE Xplore