OpenOOD: Benchmarking Generalized Out-of-Distribution Detection
This work addresses the problem of unfair comparisons and inconclusive results in OOD detection for safety-critical ML applications, though it is incremental as it focuses on benchmarking rather than new method development.
The paper tackles the lack of a unified benchmark for out-of-distribution (OOD) detection by introducing OpenOOD, a codebase that implements over 30 methods and provides a comprehensive comparison under a generalized framework, showing significant progress in the field with strong potential from preprocessing and post-hoc methods.
Out-of-distribution (OOD) detection is vital to safety-critical machine learning applications and has thus been extensively studied, with a plethora of methods developed in the literature. However, the field currently lacks a unified, strictly formulated, and comprehensive benchmark, which often results in unfair comparisons and inconclusive results. From the problem setting perspective, OOD detection is closely related to neighboring fields including anomaly detection (AD), open set recognition (OSR), and model uncertainty, since methods developed for one domain are often applicable to each other. To help the community to improve the evaluation and advance, we build a unified, well-structured codebase called OpenOOD, which implements over 30 methods developed in relevant fields and provides a comprehensive benchmark under the recently proposed generalized OOD detection framework. With a comprehensive comparison of these methods, we are gratified that the field has progressed significantly over the past few years, where both preprocessing methods and the orthogonal post-hoc methods show strong potential.