mlsanityData to 14 Oct 2026

BayesDiff: Estimating Pixel-wise Uncertainty in Diffusion via Bayesian Inference

arXiv 2310.11142 ↗First posted 17 Oct 2023cs.CV cs.LGcited by 35

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Diffusion models have impressive image generation capability, but low-quality generations still exist, and their identification remains challenging due to the lack of a proper sample-wise metric. To address this, we propose BayesDiff, a pixel-wise uncertainty estimator for generations from diffusion models based on Bayesian inference. In particular, we derive a novel uncertainty iteration principle to characterize the uncertainty dynamics in diffusion, and leverage the last-layer Laplace approximation for efficient Bayesian inference. The estimated pixel-wise uncertainty can not only be aggregated into a sample-wise metric to filter out low-fidelity images but also aids in augmenting successful generations and rectifying artifacts in failed generations in text-to-image tasks. Extensive experiments demonstrate the efficacy of BayesDiff and its promise for practical applications.

Key facts. Cited by 35 papers in the map (data through 14 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 39 of its 50 references that are in the map. First posted on arXiv 17 Oct 2023.

Builds on

39 of 50 references.

Closest ancestors · 11

Same direction, newest first

  1. 2022All are Worth Words: A ViT Backbone for Diffusion Models25 Sep
  2. 2022DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps2 Jun
  3. 2022Elucidating the Design Space of Diffusion-Based Generative Models1 Jun
  4. 2022Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models17 Jan
  5. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  6. 2021Vector Quantized Diffusion Model for Text-to-Image Synthesis29 Nov
  7. 2021Variational Diffusion Models1 Jul
  8. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  9. 2020Score-Based Generative Modeling through Stochastic Differential Equations26 Nov
  10. 2020Denoising Diffusion Implicit Models6 Oct
  11. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 28

  1. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,355
  2. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  3. 2016Improved Techniques for Training GANscited by 2,970
  4. 2018World Modelscited by 1,513
  5. 2022RePaint: Inpainting using Denoising Diffusion Probabilistic Modelscited by 1,430
  6. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
  7. 2015Generating Sentences from a Continuous Spacecited by 612
  8. 2019A Simple Baseline for Bayesian Uncertainty in Deep Learningcited by 560
  9. 2022GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generationcited by 516
  10. 2016Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithmcited by 507
  11. 2021Laplace Redux -- Effortless Bayesian Deep Learningcited by 365
  12. 2014Stochastic Gradient Hamiltonian Monte Carlocited by 343
16 more
  1. 2015Probabilistic Backpropagation for Scalable Learning of Bayesian Neural Networkscited by 295
  2. 2023Text-to-image Diffusion Models in Generative AI: A Surveycited by 265
  3. 2020Being Bayesian, Even Just a Bit, Fixes Overconfidence in ReLU Networkscited by 231
  4. 2018A Note on the Inception Scorecited by 226
  5. 2019Cyclical Stochastic Gradient MCMC for Bayesian Deep Learningcited by 203
  6. 2022ProDiff: Progressive Fast Diffusion Model For High-Quality Text-to-Speechcited by 149
  7. 2018Fast and Scalable Bayesian Deep Learning by Weight-Perturbation in Adamcited by 131
  8. 2017Noisy Natural Gradient as Variational Inferencecited by 94
  9. 2020Bayesian Deep Learning via Subnetwork Inferencecited by 87
  10. 2021Guided-TTS: A Diffusion Model for Text-to-Speech via Classifier Guidancecited by 84
  11. 2016Structured and Efficient Variational Deep Learning with Matrix Gaussian Posteriorscited by 82
  12. 2019'In-Between' Uncertainty in Bayesian Neural Networkscited by 80
  13. 2019Topic-Guided Variational Autoencoders for Text Generationcited by 44
  14. 2021LiBRe: A Practical Bayesian Approach to Adversarial Detectioncited by 40
  15. 2017A Hybrid Convolutional Variational Autoencoder for Text Generationcited by 33
  16. 2017Bayesian GANcited by 32

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 3; 2025: 6; 2026: 26.

35 papers.

Sort:

By area:

  1. 2024Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AIcited by 62
  2. 2025Generative Uncertainty in Diffusion Modelscited by 23
  3. 2024Diffusion Model Guided Sampling with Pixel-Wise Aleatoric Uncertainty Estimationcited by 20
  4. 2025Unified Uncertainty-Aware Diffusion for Multi-Agent Trajectory Modelingcited by 15
  5. 2026Self-Refining Video Samplingcited by 15
  6. 2025Temporal Score Analysis for Understanding and Correcting Diffusion Artifactscited by 9
  7. 2025Algebraic Language Models for Inverse Design of Metamaterials via Diffusion Transformerscited by 5
  8. 2024High Frequency Matters: Uncertainty Guided Image Compression with Wavelet Diffusioncited by 4
  9. 2025Trustworthy image-to-image translation: evaluating uncertainty calibration in unpaired training scenarioscited by 4
  10. 2026Flow Matching with Uncertainty Quantification and Guidancecited by 4
  11. 2026PAD-Hand: Physics-Aware Diffusion for Hand Motion Recoverycited by 4
  12. 2026Diffusion Epistemic Uncertainty with Asymmetric Learning for Diffusion-Generated Image Detectioncited by 3
23 more
  1. 2026Uncertainty in a Single Pass: A Closed-Form Identity for One-Step Flow Matchingcited by 3
  2. 2026Heteroscedastic Diffusion for Multi-Agent Trajectory Modelingcited by 3
  3. 2025UnPose: Uncertainty-Guided Diffusion Priors for Zero-Shot Pose Estimationcited by 2
  4. 2026Quantifying Epistemic Uncertainty in Diffusion Modelscited by 2
  5. 2026The Geometry of Flow-Matching Uncertainty: A Cost-free Uncertainty Proxy and Its Application in Flow-based VLA Failure Detectioncited by 2
  6. 2026When Are Two Scores Better Than One? Investigating Ensembles of Diffusion Modelscited by 1
  7. 2026DIAMOND: Directed Inference for Artifact Mitigation in Flow Matching Modelscited by 1
  8. 2026Uncertainty-Aware Distribution-to-Distribution Flow Matching for Scientific Imagingcited by 1
  9. 2026MUMINS: Metadata-conditioned Uncertainty-aware Medical Image Next-state Synthesiscited by 1
  10. 2026Inline Critic Steers Image Editingcited by 0
  11. 2026Uncertainty Quantification for Large Language Diffusion Modelscited by 0
  12. 2026Fisher-Preserving Guidance: Training-Free Manifold Constraints for Safe Diffusion Controlcited by 0
  13. 2026Efficient and Uncertainty-Aware Diffusion Framework for Offline-to-Online Reinforcement Learningcited by 0
  14. 2026Uncertainty-Calibrated Diffusion for Reliable 3D Molecular Graph Generationcited by 0
  15. 2026Uncertainty Estimation for Molecular Diffusion Modelscited by 0
  16. 2026LM-X: Explainable Vision--Language--Action Modeling via Progress, Event, and Uncertainty Predictioncited by 0
  17. 2026Is your uncertainty map wrong, or is its target? Exact diagnostics for the Tweedie diagonal, and a gradient-free alternativecited by 0
  18. 2026Matisse: Evidence-Space Reasoning for Active 3D Reconstructioncited by 0
  19. 2026Uncertainty as a Proxy for Semantic Correctness in Diffusion-Based Medical Image Synthesiscited by 0
  20. 2026TRACE: Time-Adaptive Residual Attention Control with Content-Style Decomposition for Training-Free Diffusion Style Transfercited by 0
  21. 2026Triggering Generalist Reasoning via Predictive Uncertainty for Dual-System VLAcited by 0
  22. 2026Bayesian Entropy-based Reordering for Calibrated Diffusion Language Modelscited by 0
  23. 2026Posterior Flow Matchingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023AdaDiff: Accelerating Diffusion Models through Step-Wise Adaptive Computation
  2. 2025BUFF: Bayesian Uncertainty Guided Diffusion Probabilistic Model for Single Image Super-Resolution
  3. 2026TMPDiff: Temporal Mixed-Precision for Diffusion Models
  4. 2024Diffusion Models Need Visual Priors for Image Generation
  5. 2025EVODiff: Entropy-aware Variance Optimized Diffusion Inference
  6. 2025LowDiff: Efficient Diffusion Sampling with Low-Resolution Condition
  7. 2025Generative Modeling with Bayesian Sample Inference
  8. 2025SynBoost: A Synergistic Framework for Fast Sampling of Diffusion Models