Efficiently Assemble Normalization Layers and Regularization for Federated Domain Generalization
Stats
Domain Generalization is a challenging task requiring models to perform well on unseen domains.
The proposed method outperforms existing methods in addressing the particular problem.
gPerXAN achieves average accuracies across unseen clients of 87.94% and 71.01% on PACS and Office-Home datasets, respectively.
Quotes
"Domain shift is a formidable issue in Machine Learning that causes a model to suffer from performance degradation when tested on unseen domains."
"Our proposed method outperforms other existing methods in addressing this particular problem."