SARD: Segmentation-Aware Anomaly Synthesis via Region-Constrained Diffusion with Discriminative Mask Guidance

TL;DR

SARD introduces a diffusion-based framework with region constraints and discriminative mask guidance to generate realistic, spatially precise anomalies, improving robustness in industrial anomaly detection.

Contribution

The paper presents a novel region-constrained diffusion process and a discriminative mask guidance module for enhanced anomaly synthesis.

Findings

Outperforms existing methods in segmentation accuracy

Produces higher visual quality of anomalies

Sets new state-of-the-art in pixel-level anomaly synthesis

Abstract

Synthesizing realistic and spatially precise anomalies is essential for enhancing the robustness of industrial anomaly detection systems. While recent diffusion-based methods have demonstrated strong capabilities in modeling complex defect patterns, they often struggle with spatial controllability and fail to maintain fine-grained regional fidelity. To overcome these limitations, we propose SARD (Segmentation-Aware anomaly synthesis via Region-constrained Diffusion with discriminative mask Guidance), a novel diffusion-based framework specifically designed for anomaly generation. Our approach introduces a Region-Constrained Diffusion (RCD) process that preserves the background by freezing it and selectively updating only the foreground anomaly regions during the reverse denoising phase, thereby effectively reducing background artifacts. Additionally, we incorporate a Discriminative Mask Guidance (DMG) module into the discriminator, enabling joint evaluation of both global realism and local anomaly fidelity, guided by pixel-level masks. Extensive experiments on the MVTec-AD and BTAD datasets show that SARD surpasses existing methods in segmentation accuracy and visual quality, setting a new state-of-the-art for pixel-level anomaly synthesis.