mlsanityData to 14 Oct 2026

Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models

arXiv 2212.03860 ↗First posted 7 Dec 2022cs.LG cs.CV cs.CYcited by 424

On the map: Diffusion models → Image generation, text-to-image and editing

Abstract (arXiv)

Cutting-edge diffusion models produce images with high quality and customizability, enabling them to be used for commercial art and graphic design purposes. But do diffusion models create unique works of art, or are they replicating content directly from their training sets? In this work, we study image retrieval frameworks that enable us to compare generated images with training samples and detect when content has been replicated. Applying our frameworks to diffusion models trained on multiple datasets including Oxford flowers, Celeb-A, ImageNet, and LAION, we discuss how factors such as training set size impact rates of content replication. We also identify cases where diffusion models, including the popular Stable Diffusion model, blatantly copy from their training data.

Key facts. Cited by 424 papers in the map (data through 14 Oct 2026). Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models. Builds on 55 of its 76 references that are in the map. First posted on arXiv 7 Dec 2022.

Builds on

55 of 76 references.

Closest ancestors · 1

Same direction, newest first

  1. 2022Hierarchical Text-Conditional Image Generation with CLIP Latents13 Apr

Other references · 54

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  5. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  6. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  7. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  8. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,498
  9. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  10. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  11. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  12. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,475
38 more
  1. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  2. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  3. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  4. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,884
  5. 2020Extracting Training Data from Large Language Modelscited by 2,189
  6. 2021An Empirical Study of Training Self-Supervised Vision Transformerscited by 1,616
  7. 2022RePaint: Inpainting using Denoising Diffusion Probabilistic Modelscited by 1,430
  8. 2018Deep Clustering for Unsupervised Learning of Visual Featurescited by 1,028
  9. 2017A Closer Look at Memorization in Deep Networkscited by 942
  10. 2021Membership Inference Attacks From First Principlescited by 878
  11. 2022Quantifying Memorization Across Neural Language Modelscited by 763
  12. 2017Fine-tuning CNN Image Retrieval with No Human Annotationcited by 499
  13. 2020What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimationcited by 497
  14. 2019Does Learning Require Memorization? A Short Tale about a Long Tailcited by 460
  15. 2015Deep Metric Learning via Lifted Structured Feature Embeddingcited by 449
  16. 2019Dreaming to Distill: Data-free Knowledge Transfer via DeepInversioncited by 423
  17. 2021Multimodal datasets: misogyny, pornography, and malignant stereotypescited by 327
  18. 2020Google Landmarks Dataset v2 -- A Large-Scale Benchmark for Instance-Level Recognition and Retrievalcited by 309
  19. 2022Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Modelscited by 220
  20. 2021How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Modelscited by 196
  21. 2018Revisiting Oxford and Paris: Large-Scale Image Retrieval Benchmarkingcited by 191
  22. 2022A Self-Supervised Descriptor for Image Copy Detectioncited by 176
  23. 2016Deep Image Retrieval: Learning global representations for image searchcited by 147
  24. 2022Measuring Forgetting of Memorized Training Examplescited by 118
  25. 2022VICRegL: Self-Supervised Learning of Local Visual Featurescited by 118
  26. 2021Training Vision Transformers for Image Retrievalcited by 90
  27. 2019MultiGrain: a unified image embedding for classes and instancescited by 85
  28. 2021Deep Learning for Instance Retrieval: A Surveycited by 70
  29. 2016CNN Image Retrieval Learns from BoW: Unsupervised Fine-Tuning with Hard Examplescited by 58
  30. 2020A Non-Parametric Test to Detect Data-Copying in Generative Modelscited by 52
  31. 2014Visual Instance Retrieval with Deep Convolutional Networkscited by 48
  32. 2021The 2021 Image Similarity Dataset and Challengecited by 44
  33. 2022Canary in a Coalmine: Better Membership Inference with Ensembled Adversarial Queriescited by 40
  34. 2014Multi-scale Orderless Pooling of Deep Convolutional Activation Featurescited by 38
  35. 2021ALADIN: All Layer Adaptive Instance Normalization for Fine-grained Style Similaritycited by 34
  36. 2022When do GANs replicate? On the choice of dataset sizecited by 34
  37. 2020Towards GAN Benchmarks Which Require Generalizationcited by 31
  38. 2021Self-supervised Product Quantization for Deep Unsupervised Image Retrievalcited by 28

50 most cited of 54 All 54 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 3; 2023: 88; 2024: 129; 2025: 120; 2026: 84.

424 papers.

Sort:

By area:

  1. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  2. 2023Extracting Training Data from Diffusion Modelscited by 732
  3. 2023Erasing Concepts from Diffusion Modelscited by 539
  4. 2023Unified Concept Editing in Diffusion Modelscited by 388
  5. 2023SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generationcited by 386
  6. 2023Ablating Concepts in Text-to-Image Diffusion Modelscited by 317
  7. 2024MACE: Mass Concept Erasure in Diffusion Modelscited by 307
  8. 2023Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Modelscited by 301
  9. 2023A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPTcited by 224
  10. 2023To Generate or Not? Safety-Driven Unlearned Diffusion Models Are Still Easy To Generate Unsafe Images ... For Nowcited by 210
  11. 2023Understanding and Mitigating Copying in Diffusion Modelscited by 209
  12. 2023Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion modelscited by 205
38 more
  1. 2023Generalization in diffusion models arises from geometry-adaptive harmonic representationscited by 185
  2. 2022Fake it till you make it: Learning transferable representations from synthetic ImageNet clonescited by 172
  3. 2023One-Dimensional Adapter to Rule Them All: Concepts, Diffusion Models and Erasing Applicationscited by 169
  4. 2024Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Modelscited by 168
  5. 2025HiDream-I1: A High-Efficient Image Generative Foundation Model with Sparse Diffusion Transformercited by 141
  6. 2023CADS: Unleashing the Diversity of Diffusion Models through Condition-Annealed Samplingcited by 129
  7. 2023On Memorization in Diffusion Modelscited by 125
  8. 2023Machine Learning for Synthetic Data Generation: A Reviewcited by 120
  9. 2023Receler: Reliable Concept Erasing of Text-to-Image Diffusion Models via Lightweight Eraserscited by 120
  10. 2023Foundation Models and Fair Usecited by 116
  11. 2023Ambient Diffusion: Learning Clean Distributions from Corrupted Datacited by 111
  12. 2024Measuring Style Similarity in Diffusion Modelscited by 109
  13. 2023HoloDiffusion: Training a 3D Diffusion Model using 2D Imagescited by 106
  14. 2024An analytic theory of creativity in convolutional diffusion modelscited by 104
  15. 2024Arc2Face: A Foundation Model for ID-Consistent Human Facescited by 101
  16. 2023Art and the science of generative AI: A deeper divecited by 95
  17. 2024Detecting, Explaining, and Mitigating Memorization in Diffusion Modelscited by 95
  18. 2023MusicLDM: Enhancing Novelty in Text-to-Music Generation Using Beat-Synchronous Mixup Strategiescited by 94
  19. 2024Understanding Hallucinations in Diffusion Models through Mode Interpolationcited by 91
  20. 2023Generative models improve fairness of medical classifiers under distribution shiftscited by 90
  21. 2022Position: Considerations for Differentially Private Learning with Large-Scale Public Pretrainingcited by 88
  22. 2023Differentially Private Diffusion Models Generate Useful Synthetic Imagescited by 86
  23. 2023Circumventing Concept Erasure Methods For Text-to-Image Generative Modelscited by 85
  24. 2023When Foundation Model Meets Federated Learning: Motivations, Challenges, and Future Directionscited by 84
  25. 2025Why Diffusion Models Don't Memorize: The Role of Implicit Dynamical Regularization in Trainingcited by 81
  26. 2023Particle Guidance: non-I.I.D. Diverse Sampling with Diffusion Modelscited by 71
  27. 2023On the Generalization Properties of Diffusion Modelscited by 68
  28. 2025Spot the Fake: Large Multimodal Model-Based Synthetic Image Detection with Artifact Explanationcited by 68
  29. 2023Multi-Source Diffusion Models for Simultaneous Music Generation and Separationcited by 63
  30. 2023A Reproducible Extraction of Training Images from Diffusion Modelscited by 61
  31. 2024Unveiling and Mitigating Memorization in Text-to-image Diffusion Models through Cross Attentioncited by 58
  32. 2024R.A.C.E.: Robust Adversarial Concept Erasure for Secure Text-to-Image Diffusion Modelcited by 58
  33. 2024Erasing Undesirable Influence in Diffusion Modelscited by 57
  34. 2024Breaking the Curse of Dimensionality: Diffusion Models Efficiently Learn Low-Dimensional Distributionscited by 53
  35. 2024Towards Memorization-Free Diffusion Modelscited by 50
  36. 2024Be like a Goldfish, Don't Memorize! Mitigating Memorization in Generative LLMscited by 50
  37. 2025On the Closed-Form of Flow Matching: Generalization Does Not Arise from Target Stochasticitycited by 50
  38. 2023On the De-duplication of LAION-2Bcited by 47

50 most cited of 424 All 424 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024Not Every Image is Worth a Thousand Words: Quantifying Originality in Stable Diffusion
  2. 2023Is Synthetic Data From Diffusion Models Ready for Knowledge Distillation?
  3. 2023Measuring the Success of Diffusion Models at Imitating Human Artists
  4. 2022Markup-to-Image Diffusion Models with Scheduled Sampling
  5. 2024Diffusion Models Need Visual Priors for Image Generation
  6. 2025The Cow of Rembrandt - Analyzing Artistic Prompt Interpretation in Text-to-Image Models
  7. 2025Characterizing Photorealism and Artifacts in Diffusion Model-Generated Images
  8. 2024UnlearnCanvas: Stylized Image Dataset for Enhanced Machine Unlearning Evaluation in Diffusion Models