Pith. sign in

REVIEW 3 major objections 3 minor 1 cited by

Variational Denoising Network: Toward Blind Noise Modeling and Removal

T0 review · 3 major / 3 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Blind image denoising as joint variational inference over the clean image and per-pixel noise variance, trained with an analytic evidence lower bound.

desk verdict A solid variational blind-denoising paper whose real-benchmark claims are overstated by mixed training conditions; the core method and synthetic generalization tests hold up. read the letter →

arxiv 1908.11314 v5 pith:PE724UYI submitted 2019-08-29 cs.CV

Zongsheng Yue , Hongwei Yong , Qian Zhao , Lei Zhang , Deyu Meng This is my paper · ORCID
classification cs.CV
keywords variationalinferenceblindimagedenoisingnoiseestimationnon-i.i.d.deeplearningBayesianmodelingevidencelowerboundrestoration
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Blind image denoising is usually split into noise-level estimation and cleaning, or handled by networks that memorize training noise. This paper proposes a single variational inference framework, VDN, that treats both the clean image and a per-pixel noise-variance map as latent variables conditioned on the noisy input. An approximate posterior parameterized by two networks is trained by maximizing a variational lower bound whose three terms are analytic, so no reparameterization trick is needed. On synthetic spatially varying Gaussian noise, additive white Gaussian noise, and the real SIDD and DND benchmarks, the paper reports higher PSNR than comparison methods, including a 39.28 dB versus 38.68 dB advantage over CBDNet on SIDD validation.

What carries the argument

The engine is the analytic variational lower bound built from conjugate priors: a Gaussian prior on the latent clean image (Eq. 2) and an inverse-Gamma prior on each pixel's noise variance (Eq. 3). With the factorized posterior (Eq. 5), the evidence lower bound splits into a likelihood term involving $(y_i-\mu_i)^2 + m_i^2$ divided by the predicted inverse-Gamma scale, plus two KL-divergence terms. Because every integral is closed-form, gradients flow to both networks by ordinary backpropagation, and the S-Net can be trained without a separate noise-level label. This coupling is what lets the model infer a spatially varying noise map from a single image.

What would settle it

Generate test images with known per-pixel variance maps under spatially correlated noise (for example, apply a low-pass filter to a white-noise variance field, or simulate row-wise/banding noise), then compare the S-Net's predicted variance map to the ground truth; a large mismatch, or a clear PSNR drop versus methods that model correlation, would refute the claim that VDN generalizes to unseen non-i.i.d. noise.

Watch

Extended reading notes

Core claim

The paper's central claim is that blind denoising can be solved as variational inference over the latent clean image $z$ and per-pixel noise variances $\sigma^2$, with an approximate posterior $q(z,\sigma^2|y)=q(z|y)q(\sigma^2|y)$. A denoising network (D-Net) predicts the Gaussian parameters of $q(z|y)$, and a $\sigma$ network (S-Net) predicts the inverse-Gamma parameters of $q(\sigma^2|y)$, so the model outputs a complete generative description of the noisy image rather than a single cleaned image. Training minimizes the negative evidence lower bound, whose likelihood term couples the two networks through the predicted variances; at test time the D-Net mean is the denoised image and $\beta/(\alpha+1)$ is the estimated per-pixel variance map. The paper reports that this joint inference outperforms previous blind denoising methods on synthetic non-i.i.d. Gaussian noise, on additive white Gaussian noise, and on the real SIDD and DND benchmarks.

Load-bearing premise

The load-bearing assumption is that real noise is independent zero-mean Gaussian at each pixel, with pixel-specific variance (Eq. 1); if real camera noise is too spatially correlated or signal-dependent for this approximation, the estimated noise maps and denoised outputs will degrade.

Editorial extensions

If this is right

  • A single VDN model can be trained once on noisy/clean pairs and then, on any test image, it produces both a denoised output and a per-pixel noise variance map without a separate noise-level estimator.
  • As $\varepsilon_0^2$ in the clean-image prior approaches zero, the objective degenerates to the MSE loss used by conventional networks, giving a Bayesian explanation for why those networks overfit training noise.
  • The paper states the framework can be carried over to other restoration tasks, such as super-resolution and deblurring, by replacing the data-likelihood term.
  • On the DND benchmark the reported 39.38 dB PSNR exceeds the compared blind denoising methods, indicating that the non-i.i.d. noise model transfers to real camera noise.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Inference: if the S-Net variance maps are reliable, they could serve as per-pixel confidence for downstream tasks such as multi-image fusion or uncertainty-aware vision, an application the paper does not develop.
  • Inference: the independent-Gaussian likelihood is the main constraint, so noise with strong spatial correlation (row-wise noise, demosaicking artifacts) is a likely failure mode that would show up as biased variance maps.
  • Inference: testing S-Net under a supervised loss on known synthetic variance maps, compared with the unsupervised variational objective, would isolate how much of the gain comes from the Bayesian coupling rather than network capacity.
  • Inference: because the prior on the clean image anchors to the simulated 'clean' reference $x$, results may depend on how that reference was produced; evaluating on independently measured ground truths would clarify this dependence.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 3 minor

Summary. The paper proposes a variational blind image denoising method (VDN) that treats the latent clean image z and per-pixel noise variances sigma^2 as latent variables. The likelihood is a pixel-wise independent Gaussian (Eq. 1), the prior on z is a conjugate Gaussian centered at the training 'clean' image x (Eq. 2), and the prior on each sigma_i^2 is an inverse-gamma whose mode is a locally filtered version of (y-x)^2 (Eq. 3). A mean-field variational posterior with Gaussian and inverse-gamma factors is parameterized by two networks, the D-Net and S-Net, and the negative ELBO in Eqs. (10)-(12) is minimized by backpropagation. At test time, the D-Net mean mu is the denoised image and the S-Net output beta/(alpha+1) is the estimated noise variance map. Experiments cover synthetic non-i.i.d. Gaussian noise with spatial variance maps not seen in training, AWGN, and the SIDD and DND real benchmarks, together with hyperparameter studies.

Significance. If the claims are established, VDN is a useful contribution: it provides an interpretable training objective for blind denoising, with analytic KL terms that avoid the reparameterization trick, and it performs competitively on real benchmarks. The paper has notable strengths: the ELBO algebra in Eqs. (10)-(12) is correct; the synthetic experiments use test-time noise maps different from the training map, which is a forward-looking generalization check; training code is released; and the hyperparameter analysis includes a direct MSE baseline showing that the noise-modeling terms add about 0.3 dB. The main limitations are that the real-benchmark comparison is not training-matched and the 'Bayesian lower bound' derivation is compromised by a data-dependent prior, so the headline SOTA and 'unseen noise types' claims need qualification.

major comments (3)
  1. [Section 3.1, Eq. (3), and Section 3.3, Eq. (6)] The prior in Eq. (3) is defined through xi = G((y_hat - x_hat)^2; p), which depends on the observed noisy image y (and on the training target x). With this data-dependent p(sigma^2), the decomposition log p(y) = L + KL in Eq. (6) is not a standard ELBO for a fixed generative model: p(sigma^2) is not a prior in the model p(y,z,sigma^2), and p(y) in Eq. (6) is not the marginal likelihood of that model. The paper should either define a hierarchical model that generates xi, or state that xi is a per-example hyperparameter and treat Eq. (13) as an empirical-Bayes/regularized objective rather than a true variational lower bound. As written, the central 'full Bayesian framework' claim is not supported.
  2. [Section 4.3, Tables 3 and 4] The real-benchmark comparisons do not hold training data fixed. In Table 3, the SIDD Benchmark column reports VDN trained on SIDD Medium against DnCNN-B and CBDNet numbers from models trained on other datasets; the text acknowledges this and retrains those baselines only for the validation split. Table 4 similarly compares VDN (trained on SIDD Medium) on DND with CBDNet, FFDNet, and DnCNN-B models that were not retrained on SIDD. The reported margins (39.23 vs 33.28 on SIDD Benchmark; 39.38 vs 38.06 on DND) therefore partly reflect training-data differences rather than the variational method. The validation control with retrained baselines is good, but the same matched-training comparison should be provided for DND, or the benchmark claims should be explicitly limited to matched training conditions.
  3. [Section 1 (contribution 2) and Section 4.2] The claim that VDN removes 'noise types that have never appeared in training data' is stronger than the evidence. The synthetic experiments vary only the spatial variance map M in Eq. (14) across Cases 1-3, while the noise distribution remains pixel-wise independent Gaussian as in Eq. (1); no experiment changes the noise family (e.g., to correlated, Poissonian, or signal-dependent noise with a different functional form). The generalization result is therefore about unseen spatial variance maps, not unseen noise types, and the wording in the abstract and contributions should be tempered accordingly.
minor comments (3)
  1. [Section 3.6] The sentence saying that for very small epsilon_0^2 the objective is 'dominated by the second term of Eq. (10)' appears to be a typo: the second term of Eq. (10) involves only alpha and beta, while the term that forces mu close to x is the first term of Eq. (11). Please correct the equation reference.
  2. [Eq. (14) and Fig. 2] Please clarify whether M in Eq. (14) is a standard-deviation map or a variance map; if n = n1 * M with n1 ~ N(0,1), the per-pixel noise variance is M_ij^2, but the text refers to M as the 'noise variance map' and to predicted sigma^2 maps interchangeably.
  3. [Figure 4 caption] The caption labels both the CBDNet panel and the VDN panel as '(d)'; the last panel should be '(f)'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the ELBO derivation is self-contained and benchmark numbers are external.

full rationale

The paper's central derivation is a standard variational lower bound (Eqs. 6-13) for the explicit Bayesian model in Eqs. (1)-(3). The objective in Eq. (13) is optimized on noisy-clean training pairs, and the reported denoising and noise-estimation results on Cases 1-3, SIDD, and DND are produced by forward passes on test inputs with ground truth or official benchmark evaluation, not by fitting the reported numbers. The prior in Eq. (2) uses the training clean image x, but that is ordinary supervised training, not a prediction that reduces to its input. The S-Net is trained to match the empirical residual variance map in the training data and then evaluated on unseen test noise maps, which is predictive rather than circular. Hyperparameters eps0 and p are indeed selected on the SIDD validation set (Tables 5-6), but this is model selection, not a derivation-level circularity. The only self-citation, reference [42] for the mode of the inverse Gamma distribution, is a standard mathematical fact also attributed to Bishop [6], and it is not load-bearing. The SIDD benchmark and DND comparisons mix training conditions (VDN is trained on SIDD Medium while some published baselines were trained elsewhere), but that is an experimental confound, not a circular step in the derivation. No equation or claim reduces by construction to its own input, so the paper warrants a non-finding.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The model introduces no new physical or conceptual entities beyond the standard latent variables z and sigma2. The only load-bearing ingredients are the Gaussian likelihood, the conjugate priors, the mean-field factorization, and the amortized network parameterization.

free parameters (3)
  • eps0 (variance of z prior) = 5e-5 (synthetic), 1e-6 (real)
    Controls trust in the simulated clean image x; tuned on SIDD validation (Table 5), affects PSNR by about 0.3-0.4 dB.
  • p (window size for noise prior filter) = 7
    Gaussian filter window size in Eq. (3); chosen by hand; Table 6 shows mild sensitivity.
  • Network architectures = D-Net: U-Net depth 4; S-Net: 5-layer DnCNN, 64 channels
    Architecture choices are not derived from the model; standard designs are adopted without an ablation.
assumptions (4)
  • domain assumption Mean-field factorization q(z,sigma2|y)=q(z|y)q(sigma2|y) (Eq. 4) is a good approximation for the true posterior.
    Load-bearing: if z and sigma2 are strongly coupled in the posterior, the ELBO is loose and the learned q may be poor. No empirical check of the factorization is provided.
  • domain assumption The Gaussian likelihood per pixel (Eq. 1) adequately models real image noise.
    Real noise is signal-dependent and spatially correlated; the paper assumes independent pixel-wise Gaussian with per-pixel variance.
  • domain assumption The training 'clean' image x is a valid proxy for the latent clean image z, with z ~ N(x, eps0^2) (Eq. 2).
    SIDD clean images are estimated, not true clean; if x is biased, the prior pulls z toward the bias.
  • standard math Standard ELBO decomposition and conjugacy calculations (Eqs. 6-12).
    Standard variational inference; the analytic expressions in the paper appear correct.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Variational Denoising Network: Toward Blind Noise Modeling and Removal." pith.science (2026). https://pith.science/paper/PE724UYI

@misc{pith2026190811314,
  author       = {Pith},
  title        = {Pith review of: Variational Denoising Network: Toward Blind Noise Modeling and Removal},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PE724UYI}},
  note         = {Machine review of arXiv:1908.11314}
}
read the original abstract

Blind image denoising is an important yet very challenging problem in computer vision due to the complicated acquisition process of real images. In this work we propose a new variational inference method, which integrates both noise estimation and image denoising into a unique Bayesian framework, for blind image denoising. Specifically, an approximate posterior, parameterized by deep neural networks, is presented by taking the intrinsic clean image and noise variances as latent variables conditioned on the input noisy image. This posterior provides explicit parametric forms for all its involved hyper-parameters, and thus can be easily implemented for blind image denoising with automatic noise estimation for the test noisy image. On one hand, as other data-driven deep learning methods, our method, namely variational denoising network (VDN), can perform denoising efficiently due to its explicit form of posterior expression. On the other hand, VDN inherits the advantages of traditional model-driven approaches, especially the good generalization capability of generative models. VDN has good interpretability and can be flexibly utilized to estimate and remove complicated non-i.i.d. noise collected in real scenarios. Comprehensive experiments are performed to substantiate the superiority of our method in blind image denoising.

Figures

Figures reproduced from arXiv: 1908.11314 by the authors.

Figure 1
Figure 1. The architecture of the proposed deep variational inference network. The red solid lines denote the forward process, and the blue dotted lines mark the gradient flow direction in the BP algorithm. where µi(y; WD) and m2 i (y; WD) are designed as the prediction functions for getting posterior parameters of latent variable z directly from y. The function is represented as a network, called denoising network or D-Net, … view at source ↗
Figure 2
Figure 2. (a) The spatially variant map M for noise generation in training data. (b1)-(d1): Three different Ms on testing data in Cases 1-3. (b2)-(d2): Correspondingly predicted Ms by our method on the testing data [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Image denoising results of a typical test image in Case 2. (a) Noisy image, (b) Groundtruth, (c) CBM3D (24.63dB), (d) DnCNN-B (27.83dB), (e) FFDNet (28.06dB), (f) VDN (28.32dB) [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Denoising results on one typical image in the validation set of SIDD. (a) Noisy image, (b) Simulated “clean” image, (c) WNNM(21.80dB), (d) DnCNN (34.48dB), (e) CBDNet (34.84dB), (d) VDN (35.50dB) [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: The noise variance map predicted by our proposed VDN on SIDD and DND benchmarks. (a1-a3): The noisy image, real noise (|y − x|) and noise variance map of one typical image of SIDD validation dataset. (b1-b2): The noisy image and predicted noise variance map of one typi…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. CDI: Blind Image Restoration Fidelity Evaluation based on Consistency with Degraded Image

    eess.IV 2025-01 conditional novelty 6.0 of 10

    CDI is a wavelet-domain fidelity metric for blind image restoration that compares restored images with the degraded input rather than the reference, plus a reference-free variant and a new subjective dataset.

Reference graph

Works this paper leans on

48 extracted references · 42 canonical work pages · cited by 1 Pith paper

  1. [1]

    Abdelrahman Abdelhamed, Stephen Lin, and Michael S. Brown. A high-quality denoising dataset for smartphone cameras. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2018

    work page 2018

  2. [2]

    Adaptive multi-column deep neural networks with application to robust image denoising

    Forest Agostinelli, Michael R Anderson, and Honglak Lee. Adaptive multi-column deep neural networks with application to robust image denoising. In Advances in Neural Information Processing Systems, pages 1493–1501, 2013

    work page 2013

  3. [3]

    K-svd: An algorithm for designing overcomplete dictionaries for sparse representation

    Michal Aharon, Michael Elad, Alfred Bruckstein, et al. K-svd: An algorithm for designing overcomplete dictionaries for sparse representation. IEEE Transactions on signal processing, 54(11):4311, 2006

    work page 2006

  4. [4]

    RENOIR - A Dataset for Real Low-Light Image Noise Reduction

    Josue Anaya and Adrian Barbu. Renoir - a dataset for real low-light noise image reduction. arXiv preprint arXiv:1409.8230, 2014

    work page Pith review arXiv 2014

  5. [5]

    Contour detection and hierarchical image segmentation

    Pablo Arbelaez, Michael Maire, Charless Fowlkes, and Jitendra Malik. Contour detection and hierarchical image segmentation. IEEE Trans. Pattern Anal. Mach. Intell., 33(5):898–916, May 2011

    work page 2011

  6. [6]

    Pattern recognition and machine learning

    Christopher M Bishop. Pattern recognition and machine learning. springer, 2006

    2006

  7. [7]

    Variational inference for dirichlet process mixtures

    David M Blei, Michael I Jordan, et al. Variational inference for dirichlet process mixtures. Bayesian analysis, 1(1):121–143, 2006

    work page 2006

  8. [8]

    Unprocessing Images for Learned Raw Denoising

    Tim Brooks, Ben Mildenhall, Tianfan Xue, Jiawen Chen, Dillon Sharlet, and Jonathan T Barron. Unprocessing images for learned raw denoising. arXiv preprint arXiv:1811.11127, 2018

    work page Pith review arXiv 2018

Show all 48 references
  1. [9]

    A non-local algorithm for image denoising

    Antoni Buades, Bartomeu Coll, and J-M Morel. A non-local algorithm for image denoising. In 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), volume 2, pages 60–65. IEEE, 2005

    2005

  2. [10]

    Image denoising: Can plain neural networks compete with bm3d? In 2012 IEEE conference on computer vision and pattern recognition, pages 2392–2399

    Harold C Burger, Christian J Schuler, and Stefan Harmeling. Image denoising: Can plain neural networks compete with bm3d? In 2012 IEEE conference on computer vision and pattern recognition, pages 2392–2399. IEEE, 2012

    2012

  3. [11]

    Dabov, A

    K. Dabov, A. Foi, V . Katkovnik, and K. Egiazarian. Image denoising by sparse 3-d transform- domain collaborative filtering. IEEE Transactions on Image Processing, 16(8):2080–2095, Aug 2007

    2007

  4. [12]

    Berg, and Li Fei-Fei

    Jia Deng, Olga Russakovsky, Jonathan Krause, Michael Bernstein, Alexander C. Berg, and Li Fei-Fei. Scalable multi-label annotation. In ACM Conference on Human Factors in Comput- ing Systems (CHI), 2014

    2014

  5. [13]

    Nonlocal image restoration with bilateral variance estimation: a low-rank approach

    Weisheng Dong, Guangming Shi, and Xin Li. Nonlocal image restoration with bilateral variance estimation: a low-rank approach. IEEE transactions on image processing, 22(2):700–711, 2013

    2013

  6. [14]

    Nonlocally centralized sparse representation for image restoration

    Weisheng Dong, Lei Zhang, Guangming Shi, and Xin Li. Nonlocally centralized sparse representation for image restoration. IEEE transactions on Image Processing, 22(4):1620–1630, 2012

    2012

  7. [15]

    Deep learning

    Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep learning. MIT press, 2016

    2016

  8. [16]

    Weighted nuclear norm min- imization with application to image denoising

    Shuhang Gu, Lei Zhang, Wangmeng Zuo, and Xiangchu Feng. Weighted nuclear norm min- imization with application to image denoising. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 2862–2869, 2014

    2014

  9. [17]

    Toward convolutional blind denoising of real photographs

    Shi Guo, Zifei Yan, Kai Zhang, Wangmeng Zuo, and Lei Zhang. Toward convolutional blind denoising of real photographs. arXiv preprint arXiv:1807.04686, 2018

    2018 arXiv

  10. [18]

    Delving deep into rectifiers: Surpassing human-level performance on imagenet classification

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Delving deep into rectifiers: Surpassing human-level performance on imagenet classification. In Proceedings of the IEEE international conference on computer vision , pages 1026–1034, 2015. 10

    2015

  11. [19]

    Natural image denoising with convolutional networks

    Viren Jain and Sebastian Seung. Natural image denoising with convolutional networks. In Advances in neural information processing systems , pages 769–776, 2009

    2009

  12. [20]

    Accurate image super-resolution using very deep convolutional networks

    Jiwon Kim, Jung Kwon Lee, and Kyoung Mu Lee. Accurate image super-resolution using very deep convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1646–1654, 2016

    2016

  13. [21]

    Kingma and Jimmy Lei Ba

    Diederik P. Kingma and Jimmy Lei Ba. Adam: A method for stochastic optimization. interna- tional conference on learning representations, 2015

    2015

  14. [22]

    Auto-encoding variational bayes

    Diederik P Kingma and Max Welling. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013

    2013 arXiv

  15. [23]

    Multiscale image blind denoising

    Marc Lebrun, Miguel Colom, and Jean-Michel Morel. Multiscale image blind denoising. IEEE Transactions on Image Processing, 24(10):3149–3161, 2015

    2015

  16. [24]

    Universal denoising networks: a novel cnn architecture for image denoising

    Stamatios Lefkimmiatis. Universal denoising networks: a novel cnn architecture for image denoising. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 3204–3213, 2018

    2018

  17. [25]

    Non-local recurrent network for image restoration

    Ding Liu, Bihan Wen, Yuchen Fan, Chen Change Loy, and Thomas S Huang. Non-local recurrent network for image restoration. In Advances in Neural Information Processing Systems, pages 1673–1682, 2018

    2018

  18. [26]

    Waterloo Exploration Database: New challenges for image quality assessment models

    Kede Ma, Zhengfang Duanmu, Qingbo Wu, Zhou Wang, Hongwei Yong, Hongliang Li, and Lei Zhang. Waterloo Exploration Database: New challenges for image quality assessment models. IEEE Transactions on Image Processing, 26(2):1004–1016, Feb. 2017

    2017

  19. [27]

    Nonlocal transform-domain filter for volumetric data denoising and reconstruction

    Matteo Maggioni, Vladimir Katkovnik, Karen Egiazarian, and Alessandro Foi. Nonlocal transform-domain filter for volumetric data denoising and reconstruction. IEEE transactions on image processing, 22(1):119–133, 2013

    2013

  20. [28]

    Sparse representation for color image restoration

    Julien Mairal, Michael Elad, and Guillermo Sapiro. Sparse representation for color image restoration. IEEE Transactions on image processing, 17(1):53–69, 2008

    2008

  21. [29]

    Image restoration using very deep convo- lutional encoder-decoder networks with symmetric skip connections

    Xiaojiao Mao, Chunhua Shen, and Yu-Bin Yang. Image restoration using very deep convo- lutional encoder-decoder networks with symmetric skip connections. In Advances in neural information processing systems, pages 2802–2810, 2016

    2016

  22. [30]

    Scale-space and edge detection using anisotropic diffusion

    Pietro Perona and Jitendra Malik. Scale-space and edge detection using anisotropic diffusion. IEEE Transactions on pattern analysis and machine intelligence , 12(7):629–639, 1990

    1990

  23. [31]

    Benchmarking denoising algorithms with real photographs

    Tobias Plotz and Stefan Roth. Benchmarking denoising algorithms with real photographs. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 1586–1595, 2017

    2017

  24. [32]

    Neural nearest neighbors networks

    Tobias Plötz and Stefan Roth. Neural nearest neighbors networks. In Advances in Neural Information Processing Systems, pages 1087–1098, 2018

    2018

  25. [33]

    Neural nearest neighbors networks

    Tobias Plötz and Stefan Roth. Neural nearest neighbors networks. In S. Bengio, H. Wallach, H. Larochelle, K. Grauman, N. Cesa-Bianchi, and R. Garnett, editors, Advances in Neural Information Processing Systems 31, pages 1087–1098. Curran Associates, Inc., 2018

    2018

  26. [34]

    U-net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention, pages 234–241. Springer, 2015

    2015

  27. [35]

    Fields of experts

    Stefan Roth and Michael J Black. Fields of experts. International Journal of Computer Vision, 82(2):205, 2009

    2009

  28. [36]

    Nonlinear total variation based noise removal algorithms

    Leonid I Rudin, Stanley Osher, and Emad Fatemi. Nonlinear total variation based noise removal algorithms. Physica D: nonlinear phenomena , 60(1-4):259–268, 1992. 11

    1992

  29. [37]

    Noise removal via bayesian wavelet coring

    Eero P Simoncelli and Edward H Adelson. Noise removal via bayesian wavelet coring. In Proceedings of 3rd IEEE International Conference on Image Processing , volume 1, pages 379–382. IEEE, 1996

    1996

  30. [38]

    Memnet: A persistent memory network for image restoration

    Ying Tai, Jian Yang, Xiaoming Liu, and Chunyan Xu. Memnet: A persistent memory network for image restoration. In Proceedings of the IEEE international conference on computer vision , pages 4539–4547, 2017

    2017

  31. [39]

    Statistical calibration of ccd imaging process

    Yanghai Tsin, Visvanathan Ramesh, and Takeo Kanade. Statistical calibration of ccd imaging process. In Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001, volume 1, pages 480–487. IEEE, 2001

    2001

  32. [40]

    Image denoising and inpainting with deep neural networks

    Junyuan Xie, Linli Xu, and Enhong Chen. Image denoising and inpainting with deep neural networks. In Advances in neural information processing systems , pages 341–349, 2012

    2012

  33. [41]

    A trilateral weighted sparse coding scheme for real-world image denoising

    Jun Xu, Lei Zhang, and David Zhang. A trilateral weighted sparse coding scheme for real-world image denoising. In The European Conference on Computer Vision (ECCV) , September 2018

    2018

  34. [42]

    Robust online matrix factorization for dynamic background subtraction

    Hongwei Yong, Deyu Meng, Wangmeng Zuo, and Lei Zhang. Robust online matrix factorization for dynamic background subtraction. IEEE transactions on pattern analysis and machine intelligence, 40(7):1726–1740, 2017

    2017

  35. [43]

    Hyperspectral image restoration under complex multi-band noises

    Zongsheng Yue, Deyu Meng, Yongqing Sun, and Qian Zhao. Hyperspectral image restoration under complex multi-band noises. Remote Sensing, 10(10):1631, 2018

    2018

  36. [44]

    Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising

    Kai Zhang, Wangmeng Zuo, Yunjin Chen, Deyu Meng, and Lei Zhang. Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising. IEEE Transactions on Image Processing, 26(7):3142–3155, 2017

    2017

  37. [45]

    Ffdnet: Toward a fast and flexible solution for cnn-based image denoising

    Kai Zhang, Wangmeng Zuo, and Lei Zhang. Ffdnet: Toward a fast and flexible solution for cnn-based image denoising. IEEE Transactions on Image Processing, 27(9):4608–4622, 2018

    2018

  38. [46]

    Residual dense network for image super-resolution

    Yulun Zhang, Yapeng Tian, Yu Kong, Bineng Zhong, and Yun Fu. Residual dense network for image super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2472–2481, 2018

    2018

  39. [47]

    Blind image denoising via dependent dirichlet process tree

    Fengyuan Zhu, Guangyong Chen, Jianye Hao, and Pheng-Ann Heng. Blind image denoising via dependent dirichlet process tree. IEEE transactions on pattern analysis and machine intelligence, 39(8):1518–1531, 2017

    2017

  40. [48]

    From noise modeling to blind image denoising

    Fengyuan Zhu, Guangyong Chen, and Pheng-Ann Heng. From noise modeling to blind image denoising. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 420–429, 2016. 12

    2016

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.