Diffusion-Based Multi-Class Normality for OOD Detection: An Application to CDP Authentication

2026-07-01

AI Summary

This work addresses the challenge of multi-class normality modeling, where a single detector must simultaneously capture multiple authentic distributions and produce anomaly scores comparable across classes—a critical requirement for anti-counterfeiting CDP authentication, where genuine samples exhibit high visual similarity but subtle printing and digitization discrepancies. The study introduces, for the first time, a class-conditional diffusion model for multi-class out-of-distribution (OOD) detection, proposing a ControlNet-based multi-class normality framework. Trained exclusively on authentic multi-class CDPs, the method identifies counterfeits via class-conditional reconstruction error and incorporates a dual-template masking mechanism to focus scoring on concealed regions, thereby reducing reliance on visible binary structures. Requiring neither counterfeit samples nor threshold calibration, the approach significantly outperforms existing generative baselines on the Indigo 1x1 Base dataset.

Abstract

Reconstruction-based generative models offer a natural framework for unsupervised out-of-distribution (OOD) detection, but multi-class normality modelling requires a single detector to capture multiple in-distribution manifolds and produce comparable anomaly scores across classes. We study this problem in copy detection pattern (CDP) authentication, where authentic and counterfeit samples are visually similar but differ in subtle printing-and-digitisation (P\&D) signatures. We propose a diffusion based multi-class normality framework in which a single class-conditional ControlNet is trained exclusively on authentic CDPs from multiple P\&D classes and detects counterfeits through reconstruction error under authentic-class conditioning. We further introduce dual template masking, which hides complementary regions of the input template and scores only withheld pixels, reducing reliance on visible binary structure. On the Indigo 1 x 1 Base dataset, the proposed method outperforms traditional and adapted generative baselines under multi-class authentic-versus-counterfeit evaluation, without using counterfeit samples for training or threshold calibration.

Problem

• out-of-distribution detection
• multi-class normality
• copy detection pattern
• anomaly scoring
• unsupervised authentication

Innovation

• diffusion model
• multi-class normality
• out-of-distribution detection
• ControlNet
• dual template masking