mlsanityData to 14 Oct 2026

PAEDID: Patch Autoencoder Based Deep Image Decomposition For Pixel-level Defective Region Segmentation

arXiv 2203.14457 ↗First posted 28 Mar 2022cs.CV cs.LGcited by 6

Not on the map. Why?

Abstract (arXiv)

Unsupervised pixel-level defective region segmentation is an important task in image-based anomaly detection for various industrial applications. The state-of-the-art methods have their own advantages and limitations: matrix-decomposition-based methods are robust to noise but lack complex background image modeling capability; representation-based methods are good at defective region localization but lack accuracy in defective region shape contour extraction; reconstruction-based methods detected defective region match well with the ground truth defective region shape contour but are noisy. To combine the best of both worlds, we present an unsupervised patch autoencoder based deep image decomposition (PAEDID) method for defective region segmentation. In the training stage, we learn the common background as a deep image prior by a patch autoencoder (PAE) network. In the inference stage, we formulate anomaly detection as an image decomposition problem with the deep image prior and domain-specific regularizations. By adopting the proposed approach, the defective regions in the image can be accurately extracted in an unsupervised fashion. We demonstrate the effectiveness of the PAEDID method in simulation studies and an industrial dataset in the case study.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 25 of its 50 references that are in the map. First posted on arXiv 28 Mar 2022.

Builds on

25 of 50 references.

References · 25

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  3. 2021Towards Total Recall in Industrial Anomaly Detectioncited by 776
  4. 2020Image Segmentation Using Deep Learning: A Surveycited by 660
  5. 2020PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localizationcited by 594
  6. 2017Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discoverycited by 566
  7. 2019Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detectioncited by 564
  8. 2018Deep Learning for Generic Object Detection: A Surveycited by 392
  9. 2019Uninformed Students: Student-Teacher Anomaly Detection with Discriminative Latent Embeddingscited by 386
  10. 2021CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flowscited by 321
  11. 2018Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoderscited by 264
  12. 2021FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flowscited by 227
13 more
  1. 2021VT-ADL: A Vision Transformer Network for Image Anomaly Detection and Localizationcited by 223
  2. 2020Deep Learning for Anomaly Detection: A Reviewcited by 159
  3. 2018Deep Autoencoding Models for Unsupervised Anomaly Segmentation in Brain MR Imagescited by 146
  4. 2019Detecting the Unexpected via Image Resynthesiscited by 137
  5. 2018Efficient GAN-Based Anomaly Detectioncited by 122
  6. 2022Anomalib: A Deep Learning Library for Anomaly Detectioncited by 75
  7. 2017Towards Deeper Understanding of Variational Autoencoding Modelscited by 56
  8. 2021Deep Metric Learning for Open World Semantic Segmentationcited by 51
  9. 2021Dense outlier detection and open-set recognition based on training with noisy negative imagescited by 17
  10. 2018Segmentation of Photovoltaic Module Cells in Uncalibrated Electroluminescence Imagescited by 9
  11. 2020Additive Tensor Decomposition Considering Structural Data Informationcited by 4
  12. 2022Compressed Smooth Sparse Decompositioncited by 3
  13. 2015Screen Content Image Segmentation Using Sparse-Smooth Decompositioncited by 2

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 2; 2025: 2.

6 papers.

  1. 2023RGI: robust GAN-inversion for mask-free image inpainting and unsupervised pixel-wise anomaly detectioncited by 12
  2. 2024Uni-3DAD: GAN-Inversion Aided Universal 3D Anomaly Detection on Model-free Productscited by 5
  3. 2024Synth4Seg -- Learning Defect Data Synthesis for Defect Segmentation using Bi-level Optimizationcited by 3
  4. 2025RDDPM: Robust Denoising Diffusion Probabilistic Model for Unsupervised Anomaly Segmentationcited by 3
  5. 2023Ano-SuPs: Multi-size anomaly detection for manufactured products by identifying suspected patchescited by 1
  6. 2025Single-Step Reconstruction-Free Anomaly Detection and Segmentation via Diffusion Modelscited by 1