Out-of-Distribution Detection Based on Total Variation Estimation

TL;DR

This paper presents TV-OOD, a new out-of-distribution detection method that uses total variation estimation to effectively distinguish in- and out-of-distribution data, improving security in machine learning applications.

Contribution

The paper introduces TV-OOD, a novel out-of-distribution detection approach leveraging total variation estimation, which outperforms existing methods across multiple datasets and models.

Findings

Consistently outperforms existing OOD detection methods.

Effective across various models and datasets.

Achieves comparable or superior results in image classification.

Abstract

This paper introduces a novel approach to securing machine learning model deployments against potential distribution shifts in practical applications, the Total Variation Out-of-Distribution (TV-OOD) detection method. Existing methods have produced satisfactory results, but TV-OOD improves upon these by leveraging the Total Variation Network Estimator to calculate each input's contribution to the overall total variation. By defining this as the total variation score, TV-OOD discriminates between in- and out-of-distribution data. The method's efficacy was tested across a range of models and datasets, consistently yielding results in image classification tasks that were either comparable or superior to those achieved by leading-edge out-of-distribution detection techniques across all evaluation metrics.