Adversarial Feature Learning under Accuracy Constraint for Domain Generalization

Kei Akuzawa, Yusuke Iwasawa, Yutaka Matsuo

Keywords: Domain Generalization, Adversarial Training, Invariant Feature Learning

TL;DR: Address the trade-off caused by the dependency of classes on domains by improving domain adversarial nets

Abstract: Learning domain-invariant representation is a dominant approach for domain generalization. However, previous methods based on domain invariance overlooked the underlying dependency of classes on domains, which is responsible for the trade-off between classification accuracy and the invariance. This study proposes a novel method {\em adversarial feature learning under accuracy constraint (AFLAC)}, which maximizes domain invariance within a range that does not interfere with accuracy. Empirical validations show that the performance of AFLAC is superior to that of baseline methods, supporting the importance of considering the dependency and the efficacy of the proposed method to overcome the problem.