mlsanityData to 14 Oct 2026

Saving Foundation Flow-Matching Priors for Inverse Problems

arXiv 2511.16520 ↗First posted 20 Nov 2025cs.LG cs.CV eess.IVcited by 0

On the map: Diffusion models → Flow matching, time series and science

Abstract (arXiv)

Foundation flow-matching (FM) models promise universal priors for solving inverse problems (IPs); yet today, they trail behind domain-specific and even untrained priors. How can we unlock their potential? We introduce FMPlug, a plug-in framework that redefines how foundation FMs are used in IPs. FMPlug combines an instance-guided, time-dependent warm-start strategy with sharp Gaussianity regularization, adding problem-specific guidance while preserving the Gaussian structures. For evaluation, we consider both simple image restoration tasks and scientific IPs with a few similar samples -- where the prohibitive cost of data collection and model training hinders the development of domain-specific generative models. Our superior experimental results confirm the effectiveness of FMPlug. Overall, FMPlug paves the way for making foundation FM models practical, reusable priors for IPs, especially scientific ones with few similar samples. More details are available at https://sun-umn.github.io/xm-plug/ .

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 37 of its 55 references that are in the map. First posted on arXiv 20 Nov 2025.

Builds on

37 of 55 references.

Closest ancestors · 7

Same direction, newest first

  1. 2025FMPlug: Plug-In Foundation Flow-Matching Priors for Inverse Problems1 Aug
  2. 2025Solving Inverse Problems with FLAIR3 Jun
  3. 2025FlowDPS: Flow-Driven Posterior Sampling for Inverse Problems11 Mar
  4. 2024Flow Matching Guide and Code9 Dec
  5. 2024Steering Rectified Flow Models in the Vector Field for Controlled Image Generation27 Nov
  6. 2024D-Flow: Differentiating through Flows for Controlled Generation21 Feb
  7. 2023Training-free Linear Image Inverses via Flows25 Sep

Other references · 30

  1. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  2. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  3. 2018Neural Ordinary Differential Equationscited by 3,779
  4. 2020Implicit Neural Representations with Periodic Activation Functionscited by 2,238
  5. 2017Deep Image Priorcited by 1,098
  6. 2019StarGAN v2: Diverse Image Synthesis for Multiple Domainscited by 1,098
  7. 2025FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Spacecited by 1,006
  8. 2024WorldSimBench: Towards Video Generation Models as World Simulatorscited by 881
  9. 2025Cosmos World Foundation Model Platform for Physical AIcited by 864
  10. 2018FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Modelscited by 594
  11. 2019Algorithm Unrolling: Interpretable, Efficient Deep Learning for Signal and Image Processingcited by 421
  12. 2019Toward Real-World Single Image Super-Resolution: A New Benchmark and A New Modelcited by 403
18 more
  1. 2020Deep Learning Techniques for Inverse Problems in Imagingcited by 291
  2. 2024One-Step Effective Diffusion Network for Real-World Image Super-Resolutioncited by 246
  3. 2024A Survey on Diffusion Models for Inverse Problemscited by 180
  4. 2024PnP-Flow: Plug-and-Play Image Restoration with Flow Matchingcited by 97
  5. 2024Pixel-level and Semantic-level Adjustable Super-resolution: A Dual-LoRA Approachcited by 97
  6. 2024Guidance with Spherical Gaussian Constraint for Conditional Diffusioncited by 71
  7. 2024DMPlug: A Plug-in Method for Solving Inverse Problems with Diffusion Modelscited by 61
  8. 2021Early Stopping for Deep Image Priorcited by 44
  9. 2022Theoretical Perspectives on Deep Learning Methods in Inverse Problemscited by 22
  10. 2021Self-Validation: Early Stopping for Single-Instance Deep Generative Priorscited by 15
  11. 2023Comparing the Robustness of Modern No-Reference Image- and Video-Quality Metrics to Adversarial Attackscited by 14
  12. 2024Learning Transferable Features for Implicit Neural Representationscited by 14
  13. 2022Blind Image Deblurring with Unknown Kernel Size and Substantial Noisecited by 10
  14. 2022Practical Phase Retrieval Using Double Deep Image Priorscited by 10
  15. 2022Few-shot Transfer Learning for Holographic Image Reconstruction using a Recurrent Neural Networkcited by 7
  16. 2025Temporal-Consistent Video Restoration with Pre-trained Diffusion Modelscited by 4
  17. 2025UGoDIT: Unsupervised Group Deep Image Prior Via Transferable Weightscited by 3
  18. 2024What is Wrong with End-to-End Learning for Phase Retrieval?cited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024DAWN-FM: Data-Aware and Noise-Informed Flow Matching for Solving Inverse Problems
  2. 2026FlowSGS: Improving Flow Matching Priors for Inverse Imaging with Stochastic Interpolants
  3. 2025PnP-CM: Consistency Models as Plug-and-Play Priors for Inverse Problems
  4. 2025InverseBench: Benchmarking Plug-and-Play Diffusion Priors for Inverse Problems in Physical Sciences
  5. 2025Latent Refinement via Flow Matching for Training-free Linear Inverse Problem Solving
  6. 2026What Do Flow-Based Inverse Solvers Approximate? A Posterior-Transport View
  7. 2026Consistency Regularised Gradient Flows for Inverse Problems
  8. 2026P-Flow: Proxy-gradient Flows for Linear Inverse Problems