This paper introduces FL-PBM, a novel client-side pre-training defense mechanism for Federated Learning to mitigate backdoor attacks. It proactively identifies and neutralizes poisoned data using PCA, GMM clustering, and adaptive blurring before local model training, significantly reducing attack success rates while maintaining high model accuracy.

This paper introduces Sign-Prioritized Federated Learning (SP-FL), a novel framework that improves wireless federated learning by prioritizing the transmission of critical gradient information, specifically gradient signs. It addresses unreliable communication in resource-constrained wireless networks by implementing an importance-aware hierarchical resource allocation strategy, demonstrating significant accuracy improvements.

This paper introduces FedGIN, a federated learning framework that uses augmentation-driven generalization to enable robust medical image segmentation across different imaging modalities (CT and MRI) without sharing sensitive patient data. It demonstrates that spatial-domain augmentations, specifically Global Intensity Nonlinear (GIN), significantly improve segmentation accuracy in cross-modmodality federated settings.

This paper proposes a secure and distributed system for lung disease diagnosis, specifically COVID-19 and Pneumonia, using a hybrid federated learning-enabled ensemble model. It combines established CNN architectures with the SWIN Transformer to enhance diagnostic accuracy and ensure patient data privacy through federated learning.