Towards Total Recall in Industrial Anomaly Detection

Code

• github.com/amazon-research/patchcore-inspection
  OfficialIn paperpytorch★ 1,198
• github.com/openvinotoolkit/anomalib
  pytorch★ 5,507
• github.com/amazon-science/patchcore-inspection
  pytorch★ 1,196
• github.com/hcw-00/PatchCore_anomaly_detection
  pytorch★ 347
• github.com/rvorias/ind_knn_ad
  pytorch★ 165
• github.com/OpenAOI/anodet
  pytorch★ 83
• github.com/Burf/tfdetection
  tf★ 56
• github.com/tbcvContributor/DeepHawkeye
  pytorch★ 39
• github.com/Ultranity/Anomaly.Paddle
  paddle★ 28
• github.com/tiskw/patchcore-ad
  pytorch★ 27

Abstract

Being able to spot defective parts is a critical component in large-scale industrial manufacturing. A particular challenge that we address in this work is the cold-start problem: fit a model using nominal (non-defective) example images only. While handcrafted solutions per class are possible, the goal is to build systems that work well simultaneously on many different tasks automatically. The best performing approaches combine embeddings from ImageNet models with an outlier detection model. In this paper, we extend on this line of work and propose PatchCore, which uses a maximally representative memory bank of nominal patch-features. PatchCore offers competitive inference times while achieving state-of-the-art performance for both detection and localization. On the challenging, widely used MVTec AD benchmark PatchCore achieves an image-level anomaly detection AUROC score of up to 99.6\%, more than halving the error compared to the next best competitor. We further report competitive results on two additional datasets and also find competitive results in the few samples regime.^* Work done during a research internship at Amazon AWS. Code: github.com/amazon-research/patchcore-inspection.

Tasks

Benchmark Results

Reproductions