Generalizing to Unseen Domains: A Survey on Domain Generalization

Domain generalization (DG), i.e., out-ofdistribution generalization, has attracted increasing interests in recent years. Domain generalization deals with a challenging setting where one or several different but related domain(s) are given, and the goal is to learn a model that can generalize to an unseen test domain. Great progress has been made over the years. This paper presents the first review of recent advances in domain generalization. First, we provide a formal definition of domain generalization and discuss several related fields. We then categorize recent algorithms into three classes: data manipulation, representation learning, and learning strategy, and present some algorithms in detail for each category. Third, we introduce the commonly used datasets and applications. Finally, we summarize existing literature and present some potential research topics for the future.