From Pixel to Mask: A Survey of Out-of-Distribution Segmentation

TL;DR

This survey reviews recent advances in out-of-distribution segmentation, emphasizing methods that localize anomalous objects at pixel level for safety-critical applications like autonomous driving.

Contribution

It categorizes current OoD segmentation approaches, analyzes recent progress, and discusses future challenges and directions in the field.

Findings

Identification of four main categories of OoD segmentation methods.

Recent advances improve pixel-level localization of OoD objects.

Discussion of challenges and future research directions.

Abstract

Out-of-distribution (OoD) detection and segmentation have attracted growing attention as concerns about AI security rise. Conventional OoD detection methods identify the existence of OoD objects but lack spatial localization, limiting their usefulness in downstream tasks. OoD segmentation addresses this limitation by localizing anomalous objects at pixel-level granularity. This capability is crucial for safety-critical applications such as autonomous driving, where perception modules must not only detect but also precisely segment OoD objects, enabling targeted control actions and enhancing overall system robustness. In this survey, we group current OoD segmentation approaches into four categories: (i) test-time OoD segmentation, (ii) outlier exposure for supervised training, (iii) reconstruction-based methods, (iv) and approaches that leverage powerful models. We systematically review recent advances in OoD segmentation for autonomous-driving scenarios, identify emerging challenges, and discuss promising future research directions.