Awesome Anomaly Papers: Benchmark
=================================

This repo lists awesome anomaly detection papers. There could be some
overlap between `Papers with Code`_ as we aim to list top-performing
papers here.

============== ==================== ================ ==================
Model          Featured in Anomalib Segmentation AUC  Classification AUC
============== ==================== ================ ==================
`PatchCore`_   ✅                   98.1              99.1
`PaDiM`_       ✅                   97.5              97.9
`STFPM`_       ✅                   97.0              95.5
`InTra`_       🔜                   96.9              95.9
`SPADE`_       🔜                   96.5              85.5
`CutPaste`_    ❌                   96.0              95.2
`FCDD - SS`_   ❌                   96.0              -
`Patch-SVDD`_  ❌                   95.7              92.1
`DFR`_         ❌                   95.5              93.8
`IGD`_         ❌                   93.0              93.4
`RotPred`_     ❌                   93.0              86.3
`Cavga Weak`_  ❌                   93.0              -
`FCDD`_        ❌                   92.0              88.0
`RIAD`_        ❌                   92.0              -
`DisAug CLR`_  ❌                   90.4              86.5
`CAVGA`_       ❌                   89.0              -
`AE-SSIM`_     🔜                   87.0              -
`CutPaste FT`_ ❌                   -                 97.1
`Gaussian AD`_ ❌                   -                 95.8
`DifferNet`_   ❌                   -                 94.9
`MOCCA`_       ❌                   87.5              88.0
============== ==================== ================ ==================

.. _Papers with Code: https://paperswithcode.com/sota/anomaly-detection-on-mvtec-ad
.. _PatchCore: https://paperswithcode.com/paper/towards-total-recall-in-industrial-anomaly
.. _PaDiM: https://paperswithcode.com/paper/padim-a-patch-distribution-modeling-framework
.. _STFPM: https://paperswithcode.com/paper/student-teacher-feature-pyramid-matching-for
.. _InTra: https://paperswithcode.com/paper/inpainting-transformer-for-anomaly-detection
.. _SPADE: https://paperswithcode.com/paper/sub-image-anomaly-detection-with-deep-pyramid
.. _CutPaste: https://paperswithcode.com/paper/cutpaste-self-supervised-learning-for-anomaly
.. _FCDD - SS: https://paperswithcode.com/paper/explainable-deep-one-class-classification
.. _Patch-SVDD: https://paperswithcode.com/paper/patch-svdd-patch-level-svdd-for-anomaly
.. _DFR: https://paperswithcode.com/paper/dfr-deep-feature-reconstruction-for
.. _IGD: https://paperswithcode.com/paper/unsupervised-anomaly-detection-and
.. _RotPred: https://paperswithcode.com/paper/learning-and-evaluating-representations-for-1
.. _Cavga Weak: https://paperswithcode.com/paper/attention-guided-anomaly-detection-and
.. _FCDD: https://paperswithcode.com/paper/explainable-deep-one-class-classification
.. _RIAD: https://paperswithcode.com/paper/reconstruction-by-inpainting-for-visual
.. _DisAug CLR: https://paperswithcode.com/paper/learning-and-evaluating-representations-for-1
.. _CAVGA: https://paperswithcode.com/paper/attention-guided-anomaly-detection-and
.. _AE-SSIM: https://paperswithcode.com/paper/mvtec-ad-a-comprehensive-real-world-dataset
.. _CutPaste FT: https://paperswithcode.com/paper/cutpaste-self-supervised-learning-for-anomaly
.. _Gaussian AD: https://paperswithcode.com/paper/modeling-the-distribution-of-normal-data-in
.. _DifferNet: https://paperswithcode.com/paper/same-same-but-differnet-semi-supervised
.. _MOCCA: https://paperswithcode.com/paper/mocca-multi-layer-one-class-classification