Learning Feature Inversion for Multi-class Anomaly Detection under General-purpose COCO-AD Benchmark

TL;DR

This paper introduces a new large-scale COCO-AD dataset for multi-class anomaly detection, proposes new evaluation metrics, and presents InvAD, a GAN inversion-based framework that enhances feature reconstruction for improved anomaly detection performance.

Contribution

The work constructs a comprehensive COCO-AD dataset, proposes practical AD-specific metrics, and develops InvAD, a novel GAN inversion-based method for multi-class anomaly detection.

Findings

InvAD improves detection accuracy on COCO-AD, MVTec AD, and VisA datasets.

New metrics provide more nuanced evaluation of anomaly detection methods.

The framework demonstrates effectiveness in a multi-class unsupervised setting.

Abstract

Anomaly detection (AD) is often focused on detecting anomaly areas for industrial quality inspection and medical lesion examination. However, due to the specific scenario targets, the data scale for AD is relatively small, and evaluation metrics are still deficient compared to classic vision tasks, such as object detection and semantic segmentation. To fill these gaps, this work first constructs a large-scale and general-purpose COCO-AD dataset by extending COCO to the AD field. This enables fair evaluation and sustainable development for different methods on this challenging benchmark. Moreover, current metrics such as AU-ROC have nearly reached saturation on simple datasets, which prevents a comprehensive evaluation of different methods. Inspired by the metrics in the segmentation field, we further propose several more practical threshold-dependent AD-specific metrics, ie, m$F_1$$^{.2}_{.8}$, mAcc$^{.2}_{.8}$, mIoU$^{.2}_{.8}$, and mIoU-max. Motivated by GAN inversion's high-quality reconstruction capability, we propose a simple but more powerful InvAD framework to achieve high-quality feature reconstruction. Our method improves the effectiveness of reconstruction-based methods on popular MVTec AD, VisA, and our newly proposed COCO-AD datasets under a multi-class unsupervised setting, where only a single detection model is trained to detect anomalies from different classes. Extensive ablation experiments have demonstrated the effectiveness of each component of our InvAD. Full codes and models are available at https://github.com/zhangzjn/ader.