English

RelativeFlow: Taming Medical Image Denoising Learning with Noisy Reference

Image and Video Processing 2026-04-20 v1 Artificial Intelligence Computer Vision and Pattern Recognition

Abstract

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 \textbf{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.

Keywords

Cite

@article{arxiv.2604.15459,
  title  = {RelativeFlow: Taming Medical Image Denoising Learning with Noisy Reference},
  author = {Yuxin Liu and Yiqing Dong and Wenxue Yu and Zhan Wu and Rongjun Ge and Yang Chen and Yuting He},
  journal= {arXiv preprint arXiv:2604.15459},
  year   = {2026}
}

Comments

Accepted by CVPR 2026

R2 v1 2026-07-01T12:13:26.776Z