Class-conditioned Domain Generalization via Wasserstein Distributional Robust Optimization
It addresses domain generalization for scenarios with large class-conditional distribution shifts, which is an incremental improvement over existing methods.
The paper tackles domain generalization with class-conditional distribution shifts by extending distributional robust optimization to optimize worst-case performance over Wasserstein balls around source distributions, resulting in better performance on unseen target domains compared to non-generalization approaches.
Given multiple source domains, domain generalization aims at learning a universal model that performs well on any unseen but related target domain. In this work, we focus on the domain generalization scenario where domain shifts occur among class-conditional distributions of different domains. Existing approaches are not sufficiently robust when the variation of conditional distributions given the same class is large. In this work, we extend the concept of distributional robust optimization to solve the class-conditional domain generalization problem. Our approach optimizes the worst-case performance of a classifier over class-conditional distributions within a Wasserstein ball centered around the barycenter of the source conditional distributions. We also propose an iterative algorithm for learning the optimal radius of the Wasserstein balls automatically. Experiments show that the proposed framework has better performance on unseen target domain than approaches without domain generalization.