mlsanityData to 14 Oct 2026

Self-Validation: Early Stopping for Single-Instance Deep Generative Priors

arXiv 2110.12271 ↗First posted 23 Oct 2021cs.CV cs.LG eess.IVcited by 15

Not on the map. Why?

Abstract (arXiv)

Recent works have shown the surprising effectiveness of deep generative models in solving numerous image reconstruction (IR) tasks, even without training data. We call these models, such as deep image prior and deep decoder, collectively as single-instance deep generative priors (SIDGPs). The successes, however, often hinge on appropriate early stopping (ES), which by far has largely been handled in an ad-hoc manner. In this paper, we propose the first principled method for ES when applying SIDGPs to IR, taking advantage of the typical bell trend of the reconstruction quality. In particular, our method is based on collaborative training and self-validation: the primal reconstruction process is monitored by a deep autoencoder, which is trained online with the historic reconstructed images and used to validate the reconstruction quality constantly. Experimentally, on several IR problems and different SIDGPs, our self-validation method is able to reliably detect near-peak performance and signal good ES points. Our code is available at https://sun-umn.github.io/Self-Validation/.

Key facts. Cited by 15 papers in the map (data through 14 Oct 2026). Builds on 32 of its 49 references that are in the map. First posted on arXiv 23 Oct 2021.

Builds on

32 of 49 references.

References · 32

  1. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,498
  2. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,877
  3. 2020Implicit Neural Representations with Periodic Activation Functionscited by 2,238
  4. 2020Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domainscited by 2,200
  5. 2017Deep Image Priorcited by 1,098
  6. 2017NIMA: Neural Image Assessmentcited by 389
  7. 2016Generative Visual Manipulation on the Natural Image Manifoldcited by 348
  8. 2017Compressed Sensing using Generative Modelscited by 339
  9. 2020Deep Learning Techniques for Inverse Problems in Imagingcited by 291
  10. 2020Exploiting Deep Generative Prior for Versatile Image Restoration and Manipulationcited by 225
  11. 2019Seeing What a GAN Cannot Generatecited by 148
  12. 2018Deep Decoder: Concise Image Representations from Untrained Non-convolutional Networkscited by 120
20 more
  1. 2018"Double-DIP": Unsupervised Image Decomposition via Coupled Deep-Image-Priorscited by 114
  2. 2019Neural Blind Deconvolution Using Deep Priorscited by 113
  3. 2018Deep Geometric Prior for Surface Reconstructioncited by 110
  4. 2018Phase Retrieval Under a Generative Priorcited by 73
  5. 2019A Bayesian Perspective on the Deep Image Priorcited by 52
  6. 2019Denoising and Regularization via Exploiting the Structural Bias of Convolutional Generatorscited by 40
  7. 2020Compressive sensing with un-trained neural networks: Gradient descent finds the smoothest approximationcited by 39
  8. 2021Rethinking Deep Image Prior for Denoisingcited by 38
  9. 2021On Measuring and Controlling the Spectral Bias of the Deep Image Priorcited by 36
  10. 2020Implicit Rank-Minimizing Autoencodercited by 35
  11. 2019Algorithmic Guarantees for Inverse Imaging with Untrained Network Priorscited by 32
  12. 2020Robust Recovery via Implicit Bias of Discrepant Learning Rates for Double Over-parameterizationcited by 26
  13. 2020Deep Phase Decoder: Self-calibrating phase microscopy with an untrained deep neural networkcited by 24
  14. 2020Machine Learning and Computational Mathematicscited by 22
  15. 2020Solving Inverse Problems with Hybrid Deep Image Priors: the challenge of preventing overfittingcited by 18
  16. 2020Inverse Problems, Deep Learning, and Symmetry Breakingcited by 8
  17. 2019One-dimensional Deep Image Prior for Time Series Inverse Problemscited by 7
  18. 2020Phase Retrieval with Holography and Untrained Priors: Tackling the Challenges of Low-Photon Nanoscale Imagingcited by 6
  19. 2019Low Shot Learning with Untrained Neural Networks for Imaging Inverse Problemscited by 2
  20. 2019Speech Denoising by Accumulating Per-Frequency Modeling Fluctuationscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 1; 2022: 4; 2023: 3; 2024: 2; 2025: 3; 2026: 2.

15 papers.

Sort:

By area:

  1. 2024DMPlug: A Plug-in Method for Solving Inverse Problems with Diffusion Modelscited by 61
  2. 2021Early Stopping for Deep Image Priorcited by 44
  3. 2022Blind Image Deblurring with Unknown Kernel Size and Substantial Noisecited by 10
  4. 2022Practical Phase Retrieval Using Double Deep Image Priorscited by 10
  5. 2022A Validation Approach to Over-parameterized Matrix and Image Recoverycited by 8
  6. 2022NCVX: A General-Purpose Optimization Solver for Constrained Machine and Deep Learningcited by 7
  7. 2025A Baseline Method for Removing Invisible Image Watermarks using Deep Image Priorcited by 7
  8. 2023Joint Demosaicing and Denoising with Double Deep Image Priorscited by 3
  9. 2025FMPlug: Plug-In Foundation Flow-Matching Priors for Inverse Problemscited by 3
  10. 2023Robust Autoencoders for Collective Corruption Removalcited by 0
  11. 2023JPEG Information Regularized Deep Image Prior for Denoisingcited by 0
  12. 2024Atmospheric Turbulence Removal with Video Sequence Deep Visual Priorscited by 0
3 more
  1. 2025Saving Foundation Flow-Matching Priors for Inverse Problemscited by 0
  2. 2026Automating Parameter Selection in Deep Image Prior for Fluorescence Microscopy Image Denoising via Similarity-Based Parameter Transfercited by 0
  3. 2026A Zero-Shot Deep Image Prior Framework for Denoising and Deconvolution in Fluorescence Microscopycited by 0