mlsanityData to 7 Oct 2026

DefectFill: Realistic Defect Generation with Inpainting Diffusion Model for Visual Inspection

arXiv 2503.13985 ↗First posted 18 Mar 2025cs.CV cs.AI cs.LGcited by 12

On the map: Diffusion models → Flow matching, time series and science

Abstract (arXiv)

Developing effective visual inspection models remains challenging due to the scarcity of defect data. While image generation models have been used to synthesize defect images, producing highly realistic defects remains difficult. We propose DefectFill, a novel method for realistic defect generation that requires only a few reference defect images. It leverages a fine-tuned inpainting diffusion model, optimized with our custom loss functions incorporating defect, object, and attention terms. It enables precise capture of detailed, localized defect features and their seamless integration into defect-free objects. Additionally, our Low-Fidelity Selection method further enhances the defect sample quality. Experiments show that DefectFill generates high-quality defect images, enabling visual inspection models to achieve state-of-the-art performance on the MVTec AD dataset.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 30 of its 40 references that are in the map. First posted on arXiv 18 Mar 2025.

Builds on

30 of 40 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023AnomalyDiffusion: Few-Shot Anomaly Image Generation with Diffusion Model10 Dec

Other references · 29

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  5. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  6. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  7. 2020Denoising Diffusion Implicit Modelscited by 9,655
  8. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  9. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  10. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,580
  11. 2022DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generationcited by 3,286
  12. 2022An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversioncited by 2,231
17 more
  1. 2018Demystifying MMD GANscited by 1,268
  2. 2022Multi-Concept Customization of Text-to-Image Diffusioncited by 1,099
  3. 2021Towards Total Recall in Industrial Anomaly Detectioncited by 772
  4. 2023MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editingcited by 722
  5. 2023Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image Diffusion Modelscited by 696
  6. 2021CutPaste: Self-Supervised Learning for Anomaly Detection and Localizationcited by 529
  7. 2021DRAEM -- A discriminatively trained reconstruction embedding for surface anomaly detectioncited by 510
  8. 2022Blended Latent Diffusioncited by 430
  9. 2022SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentationcited by 397
  10. 2019Deep Semi-Supervised Anomaly Detectioncited by 290
  11. 2021Few-shot Image Generation via Cross-domain Correspondencecited by 206
  12. 2021Defect-GAN: High-Fidelity Defect Synthesis for Automated Defect Inspectioncited by 77
  13. 2023Few-Shot Defect Image Generation via Defect-Aware Feature Manipulationcited by 66
  14. 2023RealFill: Reference-Driven Generation for Authentic Image Completioncited by 58
  15. 2022Prototypical Residual Networks for Anomaly Detection and Localizationcited by 56
  16. 2023CLiC: Concept Learning in Contextcited by 20
  17. 2024Personalized Residuals for Concept-Driven Text-to-Image Generationcited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 5; 2026: 7.

12 papers.

  1. 2025MAGIC: Few-Shot Mask-Guided Anomaly Inpainting with Prompt Perturbation, Spatially Adaptive Guidance, and Context Awarenesscited by 1
  2. 2025Unsupervised Multi-View Visual Anomaly Detection via Progressive Homography-Guided Alignmentcited by 1
  3. 2026One-to-More: High-Fidelity Training-Free Anomaly Generation with Attention Controlcited by 1
  4. 2025Advancing Metallic Surface Defect Detection via Anomaly-Guided Pretraining on a Large Industrial Datasetcited by 0
  5. 2025UniADC: A Unified Framework for Anomaly Detection and Classificationcited by 0
  6. 2025DeltaDeno: Zero-Shot Anomaly Generation via Delta-Denoising Attributioncited by 0
  7. 2026Synthesis4AD: Synthetic Anomalies are All You Need for 3D Anomaly Detectioncited by 0
  8. 2026Large-Scale Universal Defect Generation: Foundation Models and Datasetscited by 0
  9. 2026SynSur: An end-to-end generative pipeline for synthetic industrial surface defect generation and detectioncited by 0
  10. 2026OSAGEN: Object-Aware Mask Priors and Multistage Decoupled Diffusion for Industrial Anomaly Generationcited by 0
  11. 2026Spatial Attention Supervision for Defect Localization: Exploiting Ground-Truth Masks as Training Signal in Diffusion-Augmented Defect Detectioncited by 0
  12. 2026FLASH: A "Generate Once, Synthesize Many" Framework for Synthetic Anomaly Generation in Industrial Anomaly Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Few-shot Defect Image Generation based on Consistency Modeling
  2. 2024DefFiller: Mask-Conditioned Diffusion for Salient Steel Surface Defect Generation
  3. 2024Diffusion-based Image Generation for In-distribution Data Augmentation in Surface Defect Detection
  4. 2024Leveraging Latent Diffusion Models for Training-Free In-Distribution Data Augmentation for Surface Defect Detection
  5. 2024Dual-Interrelated Diffusion Model for Few-Shot Anomaly Image Generation
  6. 2022Unsupervised Visual Defect Detection with Score-Based Generative Model
  7. 2024Defect Image Sample Generation With Diffusion Prior for Steel Surface Defect Recognition
  8. 2024A Novel Approach to Industrial Defect Generation through Blended Latent Diffusion Model with Online Adaptation