mlsanityData to 14 Oct 2026

Positive-Unlabeled Diffusion Models for Preventing Sensitive Data Generation

arXiv 2503.03789 ↗First posted 5 Mar 2025cs.LG cs.AI stat.MLcited by 2

On the map: Diffusion models → Robustness, privacy and adversarial aspects

Abstract (arXiv)

Diffusion models are powerful generative models but often generate sensitive data that are unwanted by users, mainly because the unlabeled training data frequently contain such sensitive data. Since labeling all sensitive data in the large-scale unlabeled training data is impractical, we address this problem by using a small amount of labeled sensitive data. In this paper, we propose positive-unlabeled diffusion models, which prevent the generation of sensitive data using unlabeled and sensitive data. Our approach can approximate the evidence lower bound (ELBO) for normal (negative) data using only unlabeled and sensitive (positive) data. Therefore, even without labeled normal data, we can maximize the ELBO for normal data and minimize it for labeled sensitive data, ensuring the generation of only normal data. Through experiments across various datasets and settings, we demonstrated that our approach can prevent the generation of sensitive images without compromising image quality.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Robustness, privacy and adversarial aspects, one of 14 directions in Diffusion models. Builds on 34 of its 45 references that are in the map. First posted on arXiv 5 Mar 2025.

Builds on

34 of 45 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Models30 Mar
  2. 2023Erasing Concepts from Diffusion Models13 Mar

Other references · 32

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  5. 2020Denoising Diffusion Implicit Modelscited by 9,781
  6. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  7. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  9. 2022Classifier-Free Diffusion Guidancecited by 5,847
  10. 2016Overcoming catastrophic forgetting in neural networkscited by 4,787
  11. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,101
  12. 2014Deep Learning Face Attributes in the Wildcited by 3,693
20 more
  1. 2021Structured Denoising Diffusion Models in Discrete State-Spacescited by 1,719
  2. 2020DiffWave: A Versatile Diffusion Model for Audio Synthesiscited by 1,354
  3. 2017Continual Learning with Deep Generative Replaycited by 1,158
  4. 2022Diffusion-LM Improves Controllable Text Generationcited by 1,027
  5. 2020WaveGrad: Estimating Gradients for Waveform Generationcited by 622
  6. 2022DiffuSeq: Sequence to Sequence Text Generation with Diffusion Modelscited by 450
  7. 2021Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speechcited by 447
  8. 2019Deep Semi-Supervised Anomaly Detectioncited by 291
  9. 2020The Creation and Detection of Deepfakes: A Surveycited by 247
  10. 2017Positive-Unlabeled Learning with Non-Negative Risk Estimatorcited by 219
  11. 2020Media Forensics and DeepFakes: an overviewcited by 209
  12. 2023Selective Amnesia: A Continual Learning Approach to Forgetting in Deep Generative Modelscited by 177
  13. 2016Mixture Proportion Estimation via Kernel Embedding of Distributionscited by 67
  14. 2016Class-prior Estimation for Learning from Positive and Unlabeled Datacited by 46
  15. 2016Estimating the class prior and posterior from noisy positives and unlabeled datacited by 25
  16. 2020Learning from Positive and Unlabeled Data with Arbitrary Positive Shiftcited by 23
  17. 2022The Stable Artist: Steering Semantics in Diffusion Latent Spacecited by 23
  18. 2019Autoencoding Binary Classifiers for Supervised Anomaly Detectioncited by 17
  19. 2021Positive-Unlabeled Classification under Class-Prior Shift: A Prior-invariant Approach Based on Density Ratio Estimationcited by 7
  20. 2024Deep Positive-Unlabeled Anomaly Detection for Contaminated Unlabeled Datacited by 2

Built on by

Cited by year

Citing papers per year: 2025: 2.

2 papers.

  1. 2025Accessible, Realistic, and Fair Evaluation of Positive-Unlabeled Learning Algorithmscited by 1
  2. 2025Learning Robust Diffusion Models from Imprecise Supervisioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Unlearnable Examples for Diffusion Models: Protect Data from Unauthorized Exploitation
  2. 2025Training-Free Safe Denoisers for Safe Use of Diffusion Models
  3. 2023Conditional Generation from Unconditional Diffusion Models using Denoiser Representations
  4. 2026Asymptotically Exact Negative Guidance of Diffusion Models via Positive-Unlabeled Learning
  5. 2023Differentially Private Diffusion Models Generate Useful Synthetic Images
  6. 2025ReTrack: Data Unlearning in Diffusion Models through Redirecting the Denoising Trajectory
  7. 2024Erasing Undesirable Influence in Diffusion Models
  8. 2023Censored Sampling of Diffusion Models Using 3 Minutes of Human Feedback