Open‐set learning under covariate shift

Jie-Jing Shao, Xiao-Wen Yang, Lan-Zhe Guo

Abstract: Open-set learning deals with the testing distribution where there exist samples from the classes that are unseen during training. They aim to classify the seen classes and recog- nize the unseen classes. Previous studies typically assume that the marginal distribution of the seen classes is fixed across the training and testing distributions. In many real-world applications, however, there may exist covariate shift between them, i.e., the marginal distribution of seen classes may shift. We call this kind of problem as open-set learning under covariate shift, aim to robustly classify the seen classes under covariate shift and be aware of the unseen classes.We present a new open-set learning framework with covari- ate generalization based on supervised contrastive learning, called SC–OSG, inspired by the latent connection between contrastive learning and representation invariance. Specifi- cally, we theoretically justify supervised contrastive learning that could promote the con- ditional invariance of representations, a critical condition for covariate generalization. SC– OSG generates multi-source samples to promote the representation invariance and improve the covariate generalization. Based on this, we propose a detection score that is specific to the proposed training scheme. We evaluate the effectiveness of our method on several real-world datasets, on all of which we achieve competitive results with state-of-the-art methods.