Projection Regret: Reducing Background Bias for Novelty Detection via Diffusion Models

TL;DR

This paper introduces Projection Regret, a novel diffusion-model-based method for novelty detection that reduces background bias by comparing recursive projections, significantly improving detection accuracy.

Contribution

The paper proposes Projection Regret, a new technique that mitigates background bias in diffusion-model-based novelty detection through recursive projection comparison.

Findings

PR outperforms existing methods in novelty detection accuracy.

Recursive projections effectively cancel background bias.

The method demonstrates significant improvements across multiple datasets.

Abstract

Novelty detection is a fundamental task of machine learning which aims to detect abnormal ($\textit{i.e.}$ out-of-distribution (OOD)) samples. Since diffusion models have recently emerged as the de facto standard generative framework with surprising generation results, novelty detection via diffusion models has also gained much attention. Recent methods have mainly utilized the reconstruction property of in-distribution samples. However, they often suffer from detecting OOD samples that share similar background information to the in-distribution data. Based on our observation that diffusion models can \emph{project} any sample to an in-distribution sample with similar background information, we propose \emph{Projection Regret (PR)}, an efficient novelty detection method that mitigates the bias of non-semantic information. To be specific, PR computes the perceptual distance between the test image and its diffusion-based projection to detect abnormality. Since the perceptual distance often fails to capture semantic changes when the background information is dominant, we cancel out the background bias by comparing it against recursive projections. Extensive experiments demonstrate that PR outperforms the prior art of generative-model-based novelty detection methods by a significant margin.