/awesome-industrial-anomaly-detection

Paper list and datasets for industrial image anomaly detection.

Awesome Industrial Anomaly Detection

We discuss public datasets and related studies in detail. Welcome to read our paper and make comments.

Deep Industrial Image Anomaly Detection: A Survey

We will keep focusing on this field and update relevant information.

(Keywords: anomaly detection, anomaly segmentation, industrial image, defect detection)

Recent research

  • WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation [CVPR 2023]
  • SimpleNet: A Simple Network for Image Anomaly Detection and Localization [CVPR 2023][code]
  • Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detection [2023][code]
  • Set Features for Fine-grained Anomaly Detection[2023][code]
  • DiffusionAD: Denoising Diffusion for Anomaly Detection [2023]
  • SSGD: A smartphone screen glass dataset for defect detection [2023][dataset is coming soon]
  • In-painting Radiography Images for Unsupervised Anomaly Detection [CVPR 2023]
  • Diversity-Measurable Anomaly Detection [CVPR 2023]
  • PyramidFlow: High-Resolution Defect Contrastive Localization using Pyramid Normalizing Flow [CVPR 2023]
  • Multimodal Industrial Anomaly Detection via Hybrid Fusion [CVPR 2023][code]
  • Pushing the Limits of Fewshot Anomaly Detection in Industry Vision: Graphcore [ICLR 2023]
  • RGI: robust GAN-inversion for mask-free image inpainting and unsupervised pixel-wise anomaly detection [ICLR 2023]
  • Collaborative Discrepancy Optimization for Reliable Image Anomaly Localization [TII 2023][code]
  • CVPR 1st workshop on Vision-based InduStrial InspectiON[homepage][data link]

Paper Tree (Classification of representative methods)

Timeline

Paper list for industrial image anomaly detection

Related Survey, Benchmark and Framework

  • A review on computer vision based defect detection and condition assessment of concrete and asphalt civil infrastructure [2015]
  • Visual-based defect detection and classification approaches for industrial applications: a survey [2020]
  • Deep Learning for Unsupervised Anomaly Localization in Industrial Images: A Survey [TIM 2022]
  • A Survey on Unsupervised Industrial Anomaly Detection Algorithms [2022]
  • Benchmarking Unsupervised Anomaly Detection and Localization [2022]
  • IM-IAD: Industrial Image Anomaly Detection Benchmark in Manufacturing [2023]
  • A Deep Learning-based Software for Manufacturing Defect Inspection[TII 2017][code]
  • Anomalib: A Deep Learning Library for Anomaly Detection [code]

2 Unsupervised AD

2.1 Feature-Embedding-based Methods

2.1.1 Teacher-Student

  • Uninformed students: Student-teacher anomaly detection with discriminative latent embeddings [CVPR 2020]
  • Multiresolution knowledge distillation for anomaly detection [CVPR 2021]
  • Glancing at the Patch: Anomaly Localization With Global and Local Feature Comparison [CVPR 2021]
  • Reconstruction Student with Attention for Student-Teacher Pyramid Matching [2021]
  • Student-Teacher Feature Pyramid Matching for Anomaly Detection [2021][code]
  • PFM and PEFM for Image Anomaly Detection and Segmentation [CASE 2022][TII 2022][code]
  • Reconstructed Student-Teacher and Discriminative Networks for Anomaly Detection [2022]
  • Anomaly Detection via Reverse Distillation from One-Class Embedding [CVPR 2022][code]
  • Asymmetric Student-Teacher Networks for Industrial Anomaly Detection [WACV 2022][code]
  • Informative knowledge distillation for image anomaly segmentation [2022][code]

2.1.2 One-Class Classification (OCC)

  • Patch svdd: Patch-level svdd for anomaly detection and segmentation [ACCV 2020]
  • Anomaly detection using improved deep SVDD model with data structure preservation [2021]
  • A Semantic-Enhanced Method Based On Deep SVDD for Pixel-Wise Anomaly Detection [2021]
  • MOCCA: Multilayer One-Class Classification for Anomaly Detection [2021]
  • Defect Detection of Metal Nuts Applying Convolutional Neural Networks [2021]
  • Panda: Adapting pretrained features for anomaly detection and segmentation [2021]
  • Mean-shifted contrastive loss for anomaly detection [2021]
  • Learning and Evaluating Representations for Deep One-Class Classification [2020]
  • Self-supervised learning for anomaly detection with dynamic local augmentation [2021]
  • Contrastive Predictive Coding for Anomaly Detection [2021]
  • Cutpaste: Self-supervised learning for anomaly detection and localization [ICCV 2021][unofficial code]
  • Consistent estimation of the max-flow problem: Towards unsupervised image segmentation [2020]
  • MemSeg: A semi-supervised method for image surface defect detection using differences and commonalities [2022][unofficial code]
  • SimpleNet: A Simple Network for Image Anomaly Detection and Localization [CVPR 2023][code]

2.1.3 Distribution-Map

  • A Multi-Scale A Contrario method for Unsupervised Image Anomaly Detection [2021]
  • Modeling the distribution of normal data in pre-trained deep features for anomaly detection [2021]
  • Transfer Learning Gaussian Anomaly Detection by Fine-Tuning Representations [2021]
  • PEDENet: Image anomaly localization via patch embedding and density estimation [2022]
  • Unsupervised image anomaly detection and segmentation based on pre-trained feature mapping [2022]
  • Position Encoding Enhanced Feature Mapping for Image Anomaly Detection [2022][code]
  • Focus your distribution: Coarse-to-fine non-contrastive learning for anomaly detection and localization [ICME 2022]
  • Anomaly Detection of Defect using Energy of Point Pattern Features within Random Finite Set Framework [2021][code]
  • Fastflow: Unsupervised anomaly detection and localization via 2d normalizing flows [2021][unofficial code]
  • Same same but differnet: Semi-supervised defect detection with normalizing flows [WACV 2021][code]
  • Fully convolutional cross-scale-flows for image-based defect detection [WACV 2022][code]
  • Cflow-ad: Real-time unsupervised anomaly detection with localization via conditional normalizing flows [WACV 2022][code]
  • CAINNFlow: Convolutional block Attention modules and Invertible Neural Networks Flow for anomaly detection and localization tasks [2022]
  • AltUB: Alternating Training Method to Update Base Distribution of Normalizing Flow for Anomaly Detection [2022]
  • Collaborative Discrepancy Optimization for Reliable Image Anomaly Localization [TII 2023][code]
  • PyramidFlow: High-Resolution Defect Contrastive Localization using Pyramid Normalizing Flow [CVPR 2023]

2.1.4 Memory Bank

  • Sub-image anomaly detection with deep pyramid correspondences [2020]
  • Semi-orthogonal embedding for efficient unsupervised anomaly segmentation [2021]
  • Anomaly Detection Via Self-Organizing Map [2021]
  • PaDiM: A Patch Distribution Modeling Framework for Anomaly Detection and Localization [ICPR 2021][unofficial code]
  • Industrial Image Anomaly Localization Based on Gaussian Clustering of Pretrained Feature [2021]
  • Towards total recall in industrial anomaly detection[CVPR 2022][code]
  • CFA: Coupled-Hypersphere-Based Feature Adaptation for Target-Oriented Anomaly Localization[2022][code]
  • FAPM: Fast Adaptive Patch Memory for Real-time Industrial Anomaly Detection[2022]
  • N-pad: Neighboring Pixel-based Industrial Anomaly Detection [2022]
  • Image Anomaly Detection and Localization with Position and Neighborhood Information [2022]
  • Multi-scale patch-based representation learning for image anomaly detection and segmentation [2022]
  • SPot-the-Difference Self-supervised Pre-training for Anomaly Detection and Segmentation [ECCV 2022]
  • Diversity-Measurable Anomaly Detection [CVPR 2023]

2.2 Reconstruction-Based Methods

2.2.1 Autoencoder (AE)

  • Improving unsupervised defect segmentation by applying structural similarity to autoencoders [2018]
  • Unsupervised anomaly detection using style distillation [2020]
  • Unsupervised two-stage anomaly detection [2021]
  • Dfr: Deep feature reconstruction for unsupervised anomaly segmentation [Neurocomputing 2020]
  • Unsupervised anomaly segmentation via multilevel image reconstruction and adaptive attention-level transition [2021]
  • Encoding structure-texture relation with p-net for anomaly detection in retinal images [2020]
  • Improved anomaly detection by training an autoencoder with skip connections on images corrupted with stain-shaped noise [2021]
  • Unsupervised anomaly detection for surface defects with dual-siamese network [2022]
  • Divide-and-assemble: Learning block-wise memory for unsupervised anomaly detection [ICCV 2021]
  • Reconstruction from edge image combined with color and gradient difference for industrial surface anomaly detection [2022][code]
  • Spatial Contrastive Learning for Anomaly Detection and Localization [2022]
  • Superpixel masking and inpainting for self-supervised anomaly detection [BMVC 2020]
  • Iterative image inpainting with structural similarity mask for anomaly detection [2020]
  • Self-Supervised Masking for Unsupervised Anomaly Detection and Localization [2022]
  • Reconstruction by inpainting for visual anomaly detection [PR 2021]
  • Draem-a discriminatively trained reconstruction embedding for surface anomaly detection [ICCV 2021][code]
  • DSR: A dual subspace re-projection network for surface anomaly detection [ECCV 2022][code]
  • Natural Synthetic Anomalies for Self-supervised Anomaly Detection and Localization [ECCV 2022][code]
  • Self-Supervised Training with Autoencoders for Visual Anomaly Detection [2022]
  • Self-supervised predictive convolutional attentive block for anomaly detection [CVPR 2022 oral][code]
  • Self-Supervised Masked Convolutional Transformer Block for Anomaly Detection [TPAMI 2022][code]
  • Iterative energy-based projection on a normal data manifold for anomaly localization [2019]
  • Towards visually explaining variational autoencoders [2020]
  • Deep generative model using unregularized score for anomaly detection with heterogeneous complexity [2020]
  • Anomaly localization by modeling perceptual features [2020]
  • Image anomaly detection using normal data only by latent space resampling [2020]
  • In-painting Radiography Images for Unsupervised Anomaly Detection [CVPR 2023]

2.2.2 Generative Adversarial Networks (GANs)

  • Learning semantic context from normal samples for unsupervised anomaly detection [AAAI 2021]
  • Anoseg: Anomaly segmentation network using self-supervised learning [2021]
  • Omni-frequency Channel-selection Representations for Unsupervised Anomaly Detection [2022][code]

2.2.3 Transformer

  • VT-ADL: A vision transformer network for image anomaly detection and localization [ISIE 2021]
  • ADTR: Anomaly Detection Transformer with Feature Reconstruction [2022]
  • AnoViT: Unsupervised Anomaly Detection and Localization With Vision Transformer-Based Encoder-Decoder [2022]
  • HaloAE: An HaloNet based Local Transformer Auto-Encoder for Anomaly Detection and Localization [2022]
  • Inpainting transformer for anomaly detection [ICIAP 2022]
  • Masked Swin Transformer Unet for Industrial Anomaly Detection [2022]
  • Masked Transformer for image Anomaly Localization [TII 2022]

2.2.4 Diffusion Model

  • Denoising diffusion probabilistic models [2020]
  • AnoDDPM: Anomaly Detection With Denoising Diffusion Probabilistic Models Using Simplex Noise [CVPR Workshop 2022]
  • Unsupervised Visual Defect Detection with Score-Based Generative Model[2022]
  • DiffusionAD: Denoising Diffusion for Anomaly Detection [2023]

2.3 Supervised AD

  • Neural batch sampling with reinforcement learning for semi-supervised anomaly detection [ECCV 2020]
  • Explainable Deep One-Class Classification [ICLR 2020]
  • Attention guided anomaly localization in images [ECCV 2020]
  • Mixed supervision for surface-defect detection: From weakly to fully supervised learning [2021]
  • Explainable deep few-shot anomaly detection with deviation networks [2021]
  • Logit Inducing With Abnormality Capturing for Semi-Supervised Image Anomaly Detection [2022]
  • Catching Both Gray and Black Swans: Open-set Supervised Anomaly Detection [CVPR 2022]
  • Anomaly Clustering: Grouping Images into Coherent Clusters of Anomaly Types[WACV 2023]
  • Domain adaptation for automatic OLED panel defect detection using adaptive support vector data description [IJCV 2017]
  • An effective framework of automated visual surface defect detection for metal parts [2021]
  • Interleaved Deep Artifacts-Aware Attention Mechanism for Concrete Structural Defect Classification [TIP 2021]
  • Reference-based defect detection network [TIP 2021]
  • Fabric defect detection using tactile information [ICRA 2021]
  • A lightweight spatial and temporal multi-feature fusion network for defect detection [TIP 2020]
  • Detection and segmentation of manufacturing defects with convolutional neural networks and transfer learning [2018]

3 Other Research Direction

3.1 Few-Shot AD

  • WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation [CVPR 2023]
  • Learning unsupervised metaformer for anomaly detection [ICCV 2021]
  • Registration based few-shot anomaly detection [ECCV 2022 oral][code]
  • Same same but differnet: Semi-supervised defect detection with normalizing flows [(Distribution)WACV 2021]
  • Towards total recall in industrial anomaly detection [(Memory bank)CVPR 2022]
  • A hierarchical transformation-discriminating generative model for few shot anomaly detection [ICCV 2021]
  • Anomaly detection of defect using energy of point pattern features within random finite set framework [2021]
  • MAEDAY: MAE for few and zero shot AnomalY-Detection [2022]

3.2 Noisy AD

  • Trustmae: A noise-resilient defect classification framework using memory-augmented auto-encoders with trust regions [WACV 2021]
  • Self-Supervise, Refine, Repeat: Improving Unsupervised Anomaly Detection [TMLR 2021]
  • Data refinement for fully unsupervised visual inspection using pre-trained networks [2022]
  • Latent Outlier Exposure for Anomaly Detection with Contaminated Data [ICML 2022]
  • Deep one-class classification via interpolated gaussian descriptor [AAAI 2022 oral][code]
  • SoftPatch: Unsupervised Anomaly Detection with Noisy Data [NIPS 2020][code is coming soon]

3.3 Anomaly Synthetic

  • Cutpaste: Self-supervised learning for anomaly detection and localization [(OCC)ICCV 2021][unofficial code]
  • Draem-a discriminatively trained reconstruction embedding for surface anomaly detection [(Reconstruction AE)ICCV 2021][code]
  • MemSeg: A semi-supervised method for image surface defect detection using differences and commonalities [(OCC)2022][unofficial code]
  • Multistage GAN for fabric defect detection [2019]
  • Gan-based defect synthesis for anomaly detection in fabrics [2020]
  • Defect image sample generation with GAN for improving defect recognition [2020]
  • Defective samples simulation through neural style transfer for automatic surface defect segment [2020]
  • A simulation-based few samples learning method for surface defect segmentation [2020]
  • Synthetic data augmentation for surface defect detection and classification using deep learning [2020]
  • Defect Transfer GAN: Diverse Defect Synthesis for Data Augmentation [BMVC 2022]
  • Defect-GAN: High-fidelity defect synthesis for automated defect inspectio [2021]
  • EID-GAN: Generative Adversarial Nets for Extremely Imbalanced Data Augmentation[TII 2022]

3.4 3D AD

  • Anomaly detection in 3d point clouds using deep geometric descriptors [WACV 2022]
  • Back to the feature: classical 3d features are (almost) all you need for 3D anomaly detection [2022][code]
  • Anomaly Detection Requires Better Representations [2022]
  • Asymmetric Student-Teacher Networks for Industrial Anomaly Detection [WACV 2022]
  • Multimodal Industrial Anomaly Detection via Hybrid Fusion [CVPR 2023]
  • Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detection [2023][code]

3.5 Continual AD

  • Towards Continual Adaptation in Industrial Anomaly Detection [ACM MM 2022]

3.6 Uniform AD

  • A Unified Model for Multi-class Anomaly Detection [NIPS 2022]

3.7 Logical AD

  • Beyond Dents and Scratches: Logical Constraints in Unsupervised Anomaly Detection and Localization [IJCV 2022]
  • Set Features for Fine-grained Anomaly Detection[2023][code]

4 Dataset

  • A noise robust method based on completed local binary patterns for hot-rolled steel strip surface defects [2013]
  • Deep learning based steel pipe weld defect detection [2021]
  • (SDD)Severstal: Steel Defect Detection [2019]
  • Defect detection in SEM images of nanofibrous materials [2016]
  • (GDXray)GDXray: The database of X-ray images for nondestructive testing [2015]
  • Online PCB defect detector on a new PCB defect dataset [2019]
  • Fabric inspection based on the Elo rating method [2016]
  • (KolektorSDD)Segmentation-based deep-learning approach for surface-defect detection [Journal of Intelligent Manufacturing]
  • (KolektorSDD2)Mixed supervision for surface-defect detection: From weakly to fully supervised learning [Computers in Industry 2021]
  • (RSDD)A hierarchical extractor-based visual rail surface inspection system [2017]
  • (Eyecandies)The Eyecandies Dataset for Unsupervised Multimodal Anomaly Detection and Localization [ACCV 2022]
  • (MVTec AD)MVTec AD: A comprehensive real-world dataset for unsupervised anomaly detection [CVPR 2019]
  • (MVTec 3D-AD)The mvtec 3d-ad dataset for unsupervised 3d anomaly detection and localization [VISAPP 2021]
  • (MVTec LOCO-AD)Beyond Dents and Scratches: Logical Constraints in Unsupervised Anomaly Detection and Localization [IJCV 2022]
  • (MPDD)Deep learning-based defect detection of metal parts: evaluating current methods in complex conditions [2021]
  • (BTAD)VT-ADL: A vision transformer network for image anomaly detection and localization [2021]
  • (VisA)SPot-the-Difference Self-supervised Pre-training for Anomaly Detection and Segmentation [ECCV 2022]
  • (MTD)Surface defect saliency of magnetic tile [2020]
  • (DAGM)DAGM dataset [2007]
  • CVPR 1st workshop on Vision-based InduStrial InspectiON [homepage][data link]
  • SSGD: A smartphone screen glass dataset for defect detection [2023][dataset is coming soon]

BibTex Citation

If you find this paper and repository useful, please cite our paper.

@article{liu2023deep,
  title={Deep Industrial Image Anomaly Detection: A Survey},
  author={Liu, Jiaqi and Xie, Guoyang and Wang, Jingbao and Li, Shangnian and Wang, Chengjie and Zheng, Feng and Jin, Yaochu},
  journal={arXiv e-prints},
  pages={arXiv--2301},
  year={2023}
}