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

Nguyen Ngoc-Hieu, Nguyen Hung-Quang, The-Anh Ta, Thanh Nguyen-Tang, Khoa D Doan +1 more
6/23/2023
cs.LG

No confident code implementation found for this paper.

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.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

General-Purpose Multi-Modal OOD Detection Framework

83% match2023

Score Combining for Contrastive OOD Detection

81% match2025

A Closer Look at the Learnability of Out-of-Distribution (OOD) Detection

80% match2025

Class Relevance Learning For Out-of-distribution Detection

80% match2023

SR-OOD: Out-of-Distribution Detection via Sample Repairing

80% match2023

Comprehensive OOD Detection Improvements

80% match2024

Cite this paper

@article{ngochieu2023cosine,
  title  = {A Cosine Similarity-based Method for Out-of-Distribution Detection},
  author = {Nguyen Ngoc-Hieu and Nguyen Hung-Quang and The-Anh Ta and Thanh Nguyen-Tang and Khoa D Doan and Hoang Thanh-Tung},
  year   = {2023},
  eprint = {2306.14920},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2306.14920v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2306.14920.svg)](https://codeofpaper.com/paper/2306.14920)

Discussion