mlsanityData to 14 Oct 2026

Evaluating Unsupervised Denoising Requires Unsupervised Metrics

arXiv 2210.05553 ↗First posted 11 Oct 2022cs.CVcited by 3

Not on the map. Why?

Abstract (arXiv)

Unsupervised denoising is a crucial challenge in real-world imaging applications. Unsupervised deep-learning methods have demonstrated impressive performance on benchmarks based on synthetic noise. However, no metrics are available to evaluate these methods in an unsupervised fashion. This is highly problematic for the many practical applications where ground-truth clean images are not available. In this work, we propose two novel metrics: the unsupervised mean squared error (MSE) and the unsupervised peak signal-to-noise ratio (PSNR), which are computed using only noisy data. We provide a theoretical analysis of these metrics, showing that they are asymptotically consistent estimators of the supervised MSE and PSNR. Controlled numerical experiments with synthetic noise confirm that they provide accurate approximations in practice. We validate our approach on real-world data from two imaging modalities: videos in raw format and transmission electron microscopy. Our results demonstrate that the proposed metrics enable unsupervised evaluation of denoising methods based exclusively on noisy data.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 22 of its 59 references that are in the map. First posted on arXiv 11 Oct 2022.

Builds on

22 of 59 references.

References · 22

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  2. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 7,005
  3. 2016Densely Connected Convolutional Networkscited by 6,720
  4. 2016Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoisingcited by 1,358
  5. 2018Noise2Noise: Learning Image Restoration without Clean Datacited by 533
  6. 2018Noise2Void - Learning Denoising from Single Noisy Imagescited by 364
  7. 2019Noise2Self: Blind Denoising by Self-Supervisioncited by 276
  8. 2017Benchmarking Denoising Algorithms with Real Photographscited by 209
  9. 2021Neighbor2Neighbor: Self-Supervised Denoising from Single Noisy Imagescited by 173
  10. 2019High-Quality Self-Supervised Deep Image Denoisingcited by 161
  11. 2019Robust and interpretable blind image denoising via bias-free convolutional neural networkscited by 90
  12. 2020Supervised Raw Video Denoising with a Benchmark Dataset on Dynamic Scenescited by 77
10 more
  1. 2018Real-world Noisy Image Denoising: A New Benchmarkcited by 74
  2. 2018A Poisson-Gaussian Denoising Dataset with Real Fluorescence Microscopy Imagescited by 71
  3. 2020Noise2Same: Optimizing A Self-Supervised Bound for Image Denoisingcited by 56
  4. 2018Unsupervised Learning with Stein's Unbiased Risk Estimatorcited by 47
  5. 2020Unsupervised Deep Video Denoisingcited by 44
  6. 2018Training Deep Learning Based Denoisers without Ground Truth Datacited by 43
  7. 2021Interpretable Unsupervised Diversity Denoising and Artefact Removalcited by 26
  8. 2021Adaptive Denoising via GainTuningcited by 16
  9. 2020Deep Denoising For Scientific Discovery: A Case Study In Electron Microscopycited by 8
  10. 2021Developing and Evaluating Deep Neural Network-based Denoising for Nanoparticle TEM Images with Ultra-low Signal-to-Noisecited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 2.

3 papers.

  1. 2025SAVeD: Learning to Denoise Low-SNR Video for Improved Downstream Performancecited by 1
  2. 2026Atomic Depth Estimation From Noisy Electron Microscopy Data Via Deep Learningcited by 0
  3. 2026UGD: An Unsupervised Geometric Distance for Evaluating Real-world Noisy Point Cloud Denoisingcited by 0