Mitigating Both Covariate and Conditional Shift for Domain Generalization

TL;DR

This paper introduces VAUE, a novel domain generalization method that addresses covariate and conditional shifts through visual alignment and uncertainty-guided ensemble, improving generalization to unseen domains.

Contribution

The paper proposes VAUE, combining visual alignment for covariate shift and uncertainty-guided belief ensemble for conditional shift, advancing domain generalization techniques.

Findings

Outperforms existing methods on four datasets.

Effectively reduces distribution discrepancies.

Improves generalization to unseen domains.

Abstract

Domain generalization (DG) aims to learn a model on several source domains, hoping that the model can generalize well to unseen target domains. The distribution shift between domains contains the covariate shift and conditional shift, both of which the model must be able to handle for better generalizability. In this paper, a novel DG method is proposed to deal with the distribution shift via Visual Alignment and Uncertainty-guided belief Ensemble (VAUE). Specifically, for the covariate shift, a visual alignment module is designed to align the distribution of image style to a common empirical Gaussian distribution so that the covariate shift can be eliminated in the visual space. For the conditional shift, we adopt an uncertainty-guided belief ensemble strategy based on the subjective logic and Dempster-Shafer theory. The conditional distribution given a test sample is estimated by the dynamic combination of that of source domains. Comprehensive experiments are conducted to demonstrate the superior performance of the proposed method on four widely used datasets, i.e., Office-Home, VLCS, TerraIncognita, and PACS.