Towards Effective Semantic OOD Detection in Unseen Domains: A Domain Generalization Perspective

TL;DR

This paper introduces a novel framework for semantic out-of-distribution detection across unseen domains, combining domain generalization and OOD detection regularizations to handle both covariate and semantic shifts effectively.

Contribution

It proposes a new problem setting and a dual-regularization approach that improves OOD detection in unseen domains while preserving in-distribution accuracy.

Findings

Outperforms traditional domain generalization methods in OOD detection tasks.

Maintains comparable in-distribution classification accuracy.

Effective across multiple standard benchmarks.

Abstract

Two prevalent types of distributional shifts in machine learning are the covariate shift (as observed across different domains) and the semantic shift (as seen across different classes). Traditional OOD detection techniques typically address only one of these shifts. However, real-world testing environments often present a combination of both covariate and semantic shifts. In this study, we introduce a novel problem, semantic OOD detection across domains, which simultaneously addresses both distributional shifts. To this end, we introduce two regularization strategies: domain generalization regularization, which ensures semantic invariance across domains to counteract the covariate shift, and OOD detection regularization, designed to enhance OOD detection capabilities against the semantic shift through energy bounding. Through rigorous testing on three standard domain generalization benchmarks, our proposed framework showcases its superiority over conventional domain generalization approaches in terms of OOD detection performance. Moreover, it holds its ground by maintaining comparable InD classification accuracy.