mlsanityData to 13 Oct 2026

RelativeFlow: Taming Medical Image Denoising Learning with Noisy Reference

arXiv 2604.15459 ↗First posted 16 Apr 2026eess.IV cs.AI cs.CVcited by 0

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

Abstract (arXiv)

Medical image denoising (MID) lacks absolutely clean images for supervision, leading to a noisy reference problem that fundamentally limits denoising performance. Existing simulated-supervised discriminative learning (SimSDL) and simulated-supervised generative learning (SimSGL) treat noisy references as clean targets, causing suboptimal convergence or reference-biased learning, while self-supervised learning (SSL) imposes restrictive noise assumptions that are seldom satisfied in realistic MID scenarios. We propose RelativeFlow, a flow matching framework that learns from heterogeneous noisy references and drives inputs from arbitrary quality levels toward a unified high-quality target. RelativeFlow reformulates flow matching by decomposing the absolute noise-to-clean mapping into relative noisier-to-noisy mappings, and realizes this formulation through two key components: 1) consistent transport (CoT), a displacement map that constrains relative flows to be components of and progressively compose a unified absolute flow, and 2) simulation-based velocity field (SVF), which constructs a learnable velocity field using modality-specific degradation operators to support different medical imaging modalities. Extensive experiments on Computed Tomography (CT) and Magnetic Resonance (MR) denoising demonstrate that RelativeFlow significantly outperforms existing methods, taming MID with noisy references.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 28 of its 44 references that are in the map. First posted on arXiv 16 Apr 2026.

Builds on

28 of 44 references.

Closest ancestors · 1

Same direction, newest first

  1. 2022Flow Matching for Generative Modeling6 Oct

Other references · 27

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,952
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,656
  3. 2020Denoising Diffusion Implicit Modelscited by 9,718
  4. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,028
  5. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,213
  6. 2021SwinIR: Image Restoration Using Swin Transformercited by 1,876
  7. 2016Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoisingcited by 1,354
  8. 2017A Closer Look at Memorization in Deep Networkscited by 939
  9. 2020Learning from Noisy Labels with Deep Neural Networks: A Surveycited by 581
  10. 2018Noise2Noise: Learning Image Restoration without Clean Datacited by 529
  11. 2021Solving Inverse Problems in Medical Imaging with Score-Based Generative Modelscited by 519
  12. 2017FFDNet: Toward a Fast and Flexible Solution for CNN based Image Denoisingcited by 509
15 more
  1. 2018Noise2Void - Learning Denoising from Single Noisy Imagescited by 361
  2. 2019Noise2Self: Blind Denoising by Self-Supervisioncited by 275
  3. 2018Toward Convolutional Blind Denoising of Real Photographscited by 274
  4. 2017MemNet: A Persistent Memory Network for Image Restorationcited by 244
  5. 2017Low-Dose CT with a Residual Encoder-Decoder Convolutional Neural Network (RED-CNN)cited by 224
  6. 2017Low Dose CT Image Denoising Using a Generative Adversarial Network with Wasserstein Distance and Perceptual Losscited by 161
  7. 2019Variational Denoising Network: Toward Blind Noise Modeling and Removalcited by 117
  8. 2022MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusioncited by 66
  9. 2023CoreDiff: Contextual Error-Modulated Generalized Diffusion Model for Low-Dose CT Denoising and Generalizationcited by 52
  10. 2019GRDN:Grouped Residual Dense Network for Real Image Denoising and GAN-based Real-world Noise Modelingcited by 33
  11. 2020Patch2Self: Denoising Diffusion MRI with Self-Supervised Learningcited by 28
  12. 2021SDnDTI: Self-supervised deep learning-based denoising for diffusion tensor MRIcited by 16
  13. 2021Deformed2Self: Self-Supervised Denoising for Dynamic Medical Imagingcited by 15
  14. 2022CFNet: Conditional Filter Learning with Dynamic Noise Estimation for Real Image Denoisingcited by 2
  15. 2022Calibrating Data Mismatches in Deep Learning-Based Quantitative Ultrasound Using Setting Transfer Functionscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Beyond Fixed Inference: Quantitative Flow Matching for Adaptive Image Denoising
  2. 2026Noise-Robust Conditional Flow Matching: Generating Clean Samples from Noisy Datasets
  3. 2025DiffDenoise: Self-Supervised Medical Image Denoising with Conditional Diffusion Models
  4. 2026SPARE: Structural Parameter-Free Affinity Regularization for Flow Matching
  5. 2025LatentFM: A Latent Flow Matching Approach for Generative Medical Image Segmentation
  6. 2026Training Flow Matching: The Role of Weighting and Parameterization
  7. 2025MedSymmFlow: Bridging Generative Modeling and Classification in Medical Imaging through Symmetrical Flow Matching
  8. 2025Decoupled MeanFlow: Turning Flow Models into Flow Maps for Accelerated Sampling