Can multi-label classification networks know what they don't know?

TL;DR

This paper introduces JointEnergy, a novel method for out-of-distribution detection in multi-label classification that aggregates energy scores from multiple labels, significantly improving detection performance over existing methods.

Contribution

The paper proposes JointEnergy, a simple yet effective approach for OOD detection in multi-label classification, with a mathematical interpretation and state-of-the-art results.

Findings

JointEnergy reduces FPR95 by up to 10.05% on benchmarks.

It outperforms methods based on maximum-valued scores.

Demonstrates effectiveness on MS-COCO, PASCAL-VOC, and NUS-WIDE.

Abstract

Estimating out-of-distribution (OOD) uncertainty is a central challenge for safely deploying machine learning models in the open-world environment. Improved methods for OOD detection in multi-class classification have emerged, while OOD detection methods for multi-label classification remain underexplored and use rudimentary techniques. We propose JointEnergy, a simple and effective method, which estimates the OOD indicator scores by aggregating energy scores from multiple labels. We show that JointEnergy can be mathematically interpreted from a joint likelihood perspective. Our results show consistent improvement over previous methods that are based on the maximum-valued scores, which fail to capture joint information from multiple labels. We demonstrate the effectiveness of our method on three common multi-label classification benchmarks, including MS-COCO, PASCAL-VOC, and NUS-WIDE. We show that JointEnergy can reduce the FPR95 by up to 10.05% compared to the previous best baseline, establishing state-of-the-art performance.