mlsanityData to 7 Oct 2026

Stochastic Generative Plug-and-Play Priors

arXiv 2604.03603 ↗First posted 4 Apr 2026cs.CV cs.LG eess.IVcited by 3

On the map: Diffusion models → Inverse problems and image restoration

Abstract (arXiv)

Plug-and-play (PnP) methods are widely used for solving imaging inverse problems by incorporating a denoiser into optimization algorithms. Score-based diffusion models (SBDMs) have recently demonstrated strong generative performance through a denoiser trained across a wide range of noise levels. Despite their shared reliance on denoisers, it remains unclear how to systematically use SBDMs as priors within the PnP framework without relying on reverse diffusion sampling. In this paper, we establish a score-based interpretation of PnP that justifies using pretrained SBDMs directly within PnP algorithms. Building on this connection, we introduce a stochastic generative PnP (SGPnP) framework that injects noise to better leverage the expressive generative SBDM priors, thereby improving robustness in severely ill-posed inverse problems. We provide a new theory showing that this noise injection induces optimization on a Gaussian-smoothed objective and promotes escape from strict saddle points. Experiments on challenging inverse tasks, such as multi-coil MRI reconstruction and large-mask natural image inpainting, demonstrate consistent improvement over conventional PnP methods and achieve performance competitive with diffusion-based solvers.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Placed in Inverse problems and image restoration, one of 14 directions in Diffusion models. Builds on 33 of its 43 references that are in the map. First posted on arXiv 4 Apr 2026.

Builds on

33 of 43 references.

Closest ancestors · 8

Same direction, newest first

  1. 2025PnP-CM: Consistency Models as Plug-and-Play Priors for Inverse Problems25 Sep
  2. 2024A Survey on Diffusion Models for Inverse Problems30 Sep
  3. 2024Principled Probabilistic Imaging using Diffusion Models as Plug-and-Play Priors29 May
  4. 2024Provably Robust Score-Based Diffusion Posterior Sampling for Plug-and-Play Image Reconstruction25 Mar
  5. 2023Score-based Diffusion Models for Bayesian Image Reconstruction25 May
  6. 2023Denoising Diffusion Models for Plug-and-Play Image Restoration15 May
  7. 2022Diffusion Posterior Sampling for General Noisy Inverse Problems29 Sep
  8. 2022Denoising Diffusion Restoration Models27 Jan

Other references · 25

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  3. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  4. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 5,986
  5. 2020Plug-and-Play Image Restoration with Deep Denoiser Priorcited by 467
  6. 2018fastMRI: An Open Dataset and Benchmarks for Accelerated MRIcited by 464
  7. 2016The Little Engine that Could: Regularization by Denoising (RED)cited by 303
  8. 2016Plug-and-Play ADMM for Image Restoration: Fixed Point Convergence and Applicationscited by 244
  9. 2021Bayesian imaging using Plug & Play priors: when Langevin meets Tweediecited by 96
  10. 2018Escaping Saddles with Stochastic Gradientscited by 66
  11. 2023Provable Probabilistic Imaging using Score-Based Generative Priorscited by 51
  12. 2020Scalable Plug-and-Play ADMM with Convergence Guaranteescited by 49
13 more
  1. 2017"Convex Until Proven Guilty": Dimension-Free Acceleration of Gradient Descent on Non-Convex Functionscited by 46
  2. 2023Plug-and-Play split Gibbs sampler: embedding deep generative priors in Bayesian inferencecited by 41
  3. 2023Generative Plug and Play: Posterior Sampling for Inverse Problemscited by 24
  4. 2021SGD-Net: Efficient Model-Based Deep Learning with Theoretical Guaranteescited by 18
  5. 2022Restarted Nonconvex Accelerated Gradient Descent: No More Polylogarithmic Factor in the O(ε-7/4) Complexitycited by 17
  6. 2019Online Regularization by Denoising with Applications to Phase Retrievalcited by 12
  7. 2020A Fast Stochastic Plug-and-Play ADMM for Imaging Inverse Problemscited by 12
  8. 2024Regularization by denoising: Bayesian model and Langevin-within-split Gibbs samplingcited by 9
  9. 2020Async-RED: A Provably Convergent Asynchronous Block Parallel Stochastic Method using Deep Denoising Priorscited by 8
  10. 2022Parameter-free accelerated gradient descent for nonconvex minimizationcited by 7
  11. 2024Random Walks with Tweedie: A Unified View of Score-Based Diffusion Modelscited by 6
  12. 2024Plug-and-Play Priors as a Score-Based Methodcited by 5
  13. 2018Plug-In Stochastic Gradient Methodcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Multilevel Stochastic Plug-and-Play for Sparse-View CT Reconstructioncited by 0
  2. 2026Consistency Models for Fast MRI Reconstruction Using Regularization by Denoisingcited by 0
  3. 2026Convergent Plug-and-Play Image Restoration with Annealed Noise Levelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Plug-and-Play image restoration with Stochastic deNOising REgularization
  2. 2025Image Restoration via Primal Dual Hybrid Gradient and Flow Generative Model
  3. 2025Plug-and-Play Posterior Sampling for Blind Inverse Problems
  4. 2023Solving Inverse Problems with Score-Based Generative Priors learned from Noisy Data
  5. 2026Plug-and-Play Diffusion Meets ADMM: Dual-Variable Coupling for Robust Medical Image Reconstruction
  6. 2024PnP-Flow: Plug-and-Play Image Restoration with Flow Matching
  7. 2025Score-Based Turbo Message Passing for Plug-and-Play Compressive Imaging
  8. 2023Removing Structured Noise with Diffusion Models