Plug-and-Play Diffusion Meets ADMM: Dual-Variable Coupling for Robust Medical Image Reconstruction
On the map: Diffusion models → Inverse problems and image restoration
Abstract (arXiv)
Plug-and-Play diffusion prior (PnPDP) frameworks have emerged as a powerful paradigm for solving imaging inverse problems by treating pretrained generative models as modular priors. However, we identify a critical flaw in prevailing PnP solvers (e.g., based on HQS or Proximal Gradient): they function as memoryless operators, updating estimates solely based on instantaneous gradients. This lack of historical tracking inevitably leads to non-vanishing steady-state bias, where the reconstruction fails to strictly satisfy physical measurements under heavy corruption. To resolve this, we propose Dual-Coupled PnP Diffusion (DC-PnPDP), which restores the classical dual variable to provide integral feedback, progressively enforce agreement between the data-consistency and prior. However, this rigorous geometric coupling introduces a secondary challenge: the accumulated dual residuals exhibit spectrally colored, structured artifacts that violate the Additive White Gaussian Noise (AWGN) assumption of diffusion priors, causing severe hallucinations. To bridge this gap, we introduce Spectral Homogenization (SH), a frequency-domain adaptation mechanism that modulates these structured residuals into statistically compliant pseudo-AWGN inputs. This effectively aligns the solver's rigorous optimization trajectory with the denoiser's valid statistical manifold. Extensive experiments on CT and MRI reconstruction demonstrate that our approach resolves the bias-hallucination trade-off, achieving state-of-the-art fidelity with significantly accelerated convergence. The code is available at https://github.com/duchenhe/DC-PnPDP
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Inverse problems and image restoration, one of 14 directions in Diffusion models. Builds on 36 of its 46 references that are in the map. First posted on arXiv 26 Feb 2026.
Builds on
36 of 46 references.
Closest ancestors · 15
Same direction, newest first
- 2025Dual Ascent Diffusion for Inverse Problems23 May
- 2025InverseBench: Benchmarking Plug-and-Play Diffusion Priors for Inverse Problems in Physical Sciences14 Mar
- 2025Split Gibbs Discrete Diffusion Posterior Sampling3 Mar
- 2025Solving Inverse Problems using Diffusion with Iterative Colored Renoising29 Jan
- 2024SITCOM: Step-wise Triple-Consistent Diffusion Sampling for Inverse Problems6 Oct
- 2024Improving Diffusion Inverse Problem Solving with Decoupled Noise Annealing1 Jul
- 2024Principled Probabilistic Imaging using Diffusion Models as Plug-and-Play Priors29 May
- 2024DMPlug: A Plug-in Method for Solving Inverse Problems with Diffusion Models27 May
- 2024Provably Robust Score-Based Diffusion Posterior Sampling for Plug-and-Play Image Reconstruction25 Mar
- 2023Solving Inverse Problems with Latent Diffusion Models via Hard Data Consistency16 Jul
- 2023Denoising Diffusion Models for Plug-and-Play Image Restoration15 May
- 2023Decomposed Diffusion Sampler for Accelerating Large-Scale Inverse Problems10 Mar
3 more
Other references · 21
- 2020Denoising Diffusion Probabilistic Modelscited by 20,851
- 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
- 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
- 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
- 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
- 2022Scalable Diffusion Models with Transformerscited by 5,931
- 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 2,994
- 2021Solving Inverse Problems in Medical Imaging with Score-Based Generative Modelscited by 514
- 2018fastMRI: An Open Dataset and Benchmarks for Accelerated MRIcited by 464
- 2016Plug-and-Play ADMM for Image Restoration: Fixed Point Convergence and Applicationscited by 244
- 2020AbdomenCT-1K: Is Abdominal Organ Segmentation A Solved Problem?cited by 222
- 2019Plug-and-Play Methods Provably Converge with Properly Trained Denoiserscited by 197
9 more
- 2022Plug-and-Play Methods for Integrating Physical and Learned Models in Computational Imagingcited by 94
- 2022Proximal Denoiser for Convergent Plug-and-Play Optimization with Nonconvex Regularizationcited by 79
- 2024Guidance with Spherical Gaussian Constraint for Conditional Diffusioncited by 69
- 2023Plug-and-Play split Gibbs sampler: embedding deep generative priors in Bayesian inferencecited by 41
- 2024DPER: Diffusion Prior Driven Neural Representation for Limited Angle and Sparse View CT Reconstructioncited by 17
- 2020A Fast Stochastic Plug-and-Play ADMM for Imaging Inverse Problemscited by 12
- 2023Prior Mismatch and Adaptation in PnP-ADMM with a Nonconvex Convergence Analysiscited by 10
- 2024Decoupling Training-Free Guided Diffusion by ADMMcited by 10
- 2023Convergence of Nonconvex PnP-ADMM with MMSE Denoiserscited by 3
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2025Plug-and-play Diffusion Models for Image Compressive Sensing with Data Consistency Projection
- 2026Stochastic Generative Plug-and-Play Priors
- 2025Image Restoration via Primal Dual Hybrid Gradient and Flow Generative Model
- 2025Plug-and-Play Posterior Sampling for Blind Inverse Problems
- 2025Investigating the Feasibility of Patch-based Inference for Generalized Diffusion Priors in Inverse Problems for Medical Images
- 2026A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- 2025DAPS++: Rethinking Diffusion Inverse Problems with Decoupled Posterior Annealing
- 2025Diffusion models for inverse problems