Out-of-Distribution Detection Using Outlier Detection Methods

TL;DR

This paper demonstrates that outlier detection algorithms like Isolation Forest can effectively identify out-of-distribution inputs in neural networks without requiring model adaptation, offering a simple and unsupervised solution.

Contribution

It introduces an out-of-distribution detection method using outlier detection algorithms based on softmax scores, eliminating the need for neural network modifications.

Findings

Isolation Forest achieves comparable performance to specialized OOD methods

Supervised methods like Gradient Boosting further improve detection accuracy

The approach is unsupervised and model-agnostic

Abstract

Out-of-distribution detection (OOD) deals with anomalous input to neural networks. In the past, specialized methods have been proposed to reject predictions on anomalous input. Similarly, it was shown that feature extraction models in combination with outlier detection algorithms are well suited to detect anomalous input. We use outlier detection algorithms to detect anomalous input as reliable as specialized methods from the field of OOD. No neural network adaptation is required; detection is based on the model's softmax score. Our approach works unsupervised using an Isolation Forest and can be further improved by using a supervised learning method such as Gradient Boosting.