MOOD: Multi-level Out-of-distribution Detection

TL;DR

The paper introduces MOOD, a multi-level OOD detection framework that leverages intermediate classifier outputs for efficient and accurate out-of-distribution identification, reducing computation while maintaining performance.

Contribution

MOOD is a novel framework that exploits intermediate classifier outputs for dynamic, efficient OOD detection, enabling early detection of easy OOD examples and reducing inference cost.

Findings

Achieves up to 71.05% computational reduction in inference.

Maintains competitive OOD detection performance across 10 datasets.

Establishes a theoretical and empirical basis for multi-level OOD detection.

Abstract

Out-of-distribution (OOD) detection is essential to prevent anomalous inputs from causing a model to fail during deployment. While improved OOD detection methods have emerged, they often rely on the final layer outputs and require a full feedforward pass for any given input. In this paper, we propose a novel framework, multi-level out-of-distribution detection MOOD, which exploits intermediate classifier outputs for dynamic and efficient OOD inference. We explore and establish a direct relationship between the OOD data complexity and optimal exit level, and show that easy OOD examples can be effectively detected early without propagating to deeper layers. At each exit, the OOD examples can be distinguished through our proposed adjusted energy score, which is both empirically and theoretically suitable for networks with multiple classifiers. We extensively evaluate MOOD across 10 OOD datasets spanning a wide range of complexities. Experiments demonstrate that MOOD achieves up to 71.05% computational reduction in inference, while maintaining competitive OOD detection performance.