Privacy-Aligned Personalized Federated Learning with Compact Adaptation and Variable-Length Gaussian Communication
Across MNIST and CIFAR-10, the design matches or outperforms full-model private adaptation across privacy budgets and client heterogeneity, while reducing protected uplink by a factor of 2.67 at \(\varepsilon=16\) on CIFAR-10 with comparable future-client accuracy.
Yi-Lin Xu, Chun Hei Michael Shiu, Chih-Wei Ling et al.
· 0 citations