A Cosine Similarity-based Method for Out-of-Distribution Detection

TL;DR

This paper introduces Class Typical Matching (CTM), a cosine similarity-based method for out-of-distribution detection that effectively distinguishes OOD data from in-distribution data.

Contribution

The paper proposes a novel post hoc OOD detection method using cosine similarity, demonstrating superior performance over existing methods.

Findings

CTM outperforms existing OOD detection methods on multiple benchmarks.

Cosine similarity between test features and typical ID features is an effective OOD indicator.

Extensive experiments validate the effectiveness of the proposed approach.

Abstract

The ability to detect OOD data is a crucial aspect of practical machine learning applications. In this work, we show that cosine similarity between the test feature and the typical ID feature is a good indicator of OOD data. We propose Class Typical Matching (CTM), a post hoc OOD detection algorithm that uses a cosine similarity scoring function. Extensive experiments on multiple benchmarks show that CTM outperforms existing post hoc OOD detection methods.