A malicious FL server can steal private training images by encoding them into model parameters via a correlation regularizer and preserving them through segmented aggregation.
Proceedings of the IEEE86(11), 2278–2324 (1998)
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
dataset 1
citation-polarity summary
roles
dataset 1polarities
use dataset 1representative citing papers
Jellyfish enables zero-shot federated unlearning through synthetic proxy data generation, channel-restricted knowledge disentanglement, and a composite loss with repair to forget target data while retaining model utility.
Machine unlearning methods adapted to hybrid quantum models achieve effective forgetting that varies with circuit depth and entanglement, establishing initial empirical baselines for quantum-aware unlearning.
A CNN-attention model achieves 99.2% accuracy on seen MRI sites and 75.5% on unseen heterogeneous sites for motion artifact quality assessment.
citing papers explorer
-
FedCVESA: Taking Away Training Data in Federated Learning via Correlation Value Encoding and Segmented Aggregation
A malicious FL server can steal private training images by encoding them into model parameters via a correlation regularizer and preserving them through segmented aggregation.
-
Jellyfish: Zero-Shot Federated Unlearning Scheme with Knowledge Disentanglement
Jellyfish enables zero-shot federated unlearning through synthetic proxy data generation, channel-restricted knowledge disentanglement, and a composite loss with repair to forget target data while retaining model utility.
-
Machine Unlearning in the Era of Quantum Machine Learning: An Empirical Study
Machine unlearning methods adapted to hybrid quantum models achieve effective forgetting that varies with circuit depth and entanglement, establishing initial empirical baselines for quantum-aware unlearning.
-
Attention-Gated Convolutional Networks for Scanner-Agnostic Quality Assessment
A CNN-attention model achieves 99.2% accuracy on seen MRI sites and 75.5% on unseen heterogeneous sites for motion artifact quality assessment.