The DOI matches its Crossref record; 74 of the 81 references it cites were found in a registry or an index; 2 of the 26 sources its claims cite were checked for support.
Learning Deep CNN Denoiser Prior for Image Restoration
Open accessAbstract
Model-based optimization methods and discriminative learning methods have been the two dominant strategies for solving various inverse problems in low-level vision. Typically, those two kinds of methods have their respective merits and drawbacks, e.g., model-based optimization methods are flexible for handling different inverse problems but are usually time-consuming with sophisticated priors for the purpose of good performance, in the meanwhile, discriminative learning methods have fast testing speed but their application range is greatly restricted by the specialized task. Recent works have revealed that, with the aid of variable splitting techniques, denoiser prior can be plugged in as a modular part of model-based optimization methods to solve other inverse problems (e.g., deblurring). Such an integration induces considerable advantage when the denoiser is obtained via discriminative learning. However, the study of integration with fast discriminative denoiser prior is still lacking. To this end, this paper aims to train a set of fast and effective CNN (convolutional neural network) denoisers and integrate them into model-based optimization method to solve other inverse problems. Experimental results demonstrate that the learned set of denoisers can not only achieve promising Gaussian denoising results but also can be used as prior to deliver good performance for various low-level vision applications.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
The paper trains a set of fast CNN Gaussian denoisers and uses them as modular priors inside a half-quadratic splitting optimization framework to solve image deblurring and super-resolution. Experiments show competitive denoising accuracy and flexible restoration performance across tasks without task-specific retraining.
Key findings
- The proposed CNN denoiser improves gray-image denoising on BSD68 over BM3D, WNNM, TNRD, and MLP: at σ=15 it reaches 31.63 dB, at σ=25 it reaches 29.15 dB, and at σ=50 it reaches 26.19 dB.
- For color denoising on BSD68, the proposed denoiser consistently beats CBM3D, e.g. 40.36 vs. 40.24 dB at noise level 5, 31.16 vs. 30.71 dB at 25, and 27.86 vs. 27.38 dB at 50.
- The learned denoisers can be inserted into a model-based optimization method to handle deblurring and single-image super-resolution; for example, on the Leaves deblurring case with Gaussian blur σ=1.6 and noise level 2, the proposed method reports 29.78 dB versus 26.95 dB for IDDBM3D and 27.50 dB for NCSR.
- Runtime is competitive: for a 256×256 image with noise level 25, the 24-feature-map version takes 0.10 s on CPU and 0.006 s on GPU, while BM3D takes 0.66 s on CPU and TNRD takes 0.47 s on CPU / 0.010 s on GPU.
Methods
- The authors design a seven-layer CNN denoiser using dilated convolutions with dilation factors 1, 2, 3, 4, 3, 2, 1; middle layers use 64 feature maps, and the effective receptive field is 33×33.
- Training uses 400 BSD images, 400 ImageNet validation images, and 4,744 Waterloo Exploration Database images, cropped into 35×35 patches with N = 256×4,000 patches; Adam is used with residual learning and batch normalization.
- A set of 25 denoisers is trained for gray images and 25 for color images over noise levels from 0 to 50 in steps of 2; these denoisers are integrated into a half-quadratic splitting scheme, where the regularization subproblem becomes a Gaussian denoising step.
Limitations
- The paper states that future work includes reducing the number of CNN denoisers and iterations, extending the framework to inpainting and blind deblurring, and using multiple complementary priors.
- The experimental evaluation is limited to denoising, non-blind deblurring, and super-resolution; inpainting and blind deblurring are not tested.
- The authors note that comparisons with SRCNN and VDSR for non-bicubic super-resolution are unfair because those methods were trained for bicubic degradation, which is evident from the reported setup rather than presented as a limitation of the proposed method.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
2 of 26 cited sources checked · 1 supported · 1 flagged · accuracy not yet measured
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“Note that SRCNN and VDSR are trained with bicubic blur kernel, thus it is unfair to use their models to super-resolve the low-resolution image with Gaussian kernel.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.3. Single Image Super-Resolution
What the sources say
- Disputedcites [21] 21. C. Dong, C. C. Loy, K. He, and X. Tang. Image super-resolution using deep convolutional networks. IEEE transactions on Pattern Analysis and Machine Intelligence, 38(2):295-307, 2016. 2, 7 sourceExperimental — accuracy not yet measured
Section 3.3 (p. 5) says SRCNN training samples are blurred with a Gaussian kernel; bicubic interpolation is a later upscaling step, not the blur kernel. Section 4.4.1 (p. 9) also describes training a separate SRCNN network with a different Gaussian kernel for comparison. The paper does not address VDSR's training. (LLM, conf=0.87)
Checked by a reasoning model — may be wrong.
- Not checked — no full text of the sourcecites [35] 35. J. Kim, J. K. Lee, and K. M. Lee. Accurate image super- resolution using very deep convolutional networks. In IEEE Conference on Computer Vision and Pattern Recognition, pages 1646-1654, 2016. 4, 7 source
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Disputed
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Experimental results demonstrate that the learned set of denoisers can not only achieve promising Gaussian denoising results but also can be used as prior to deliver good performance for various low-level vision applications.”
- Also stated in: § 1. Introduction · boosted
“Extensive experiments on classical IR problems, including deblurring and super-resolution, have demonstrated the merits of integrating flexible model-based optimization methods and fast CNN-based discriminative learning methods.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Abstract
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Also stated in: § 1. Introduction · boosted
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“we empirically find it also can enable fast transfer from one model to another with different noise level.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Using Batch Normalization and Residual Learning to Accelerate Training.
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesThe paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“However, if the size of the training patch is smaller than the receptive field, the performance would decrease.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Using Batch Normalization and Residual Learning to Accelerate Training.
Strength:hedged· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, the proposed CNN denoiser can have a PSNR gain of about 0.2dB over those three methods.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
Load-bearingStrength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesThe paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“We empirically find that using training samples with small size can help avoid boundary artifacts.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Using Batch Normalization and Residual Learning to Accelerate Training.
Strength:plain· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We can see that the proposed denoiser is very competitive in both CPU and GPU implementation.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesThe paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Due to the iterative scheme, it turns out the noise level range of [0, 50] is enough to handle various image restoration problems.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Using Batch Normalization and Residual Learning to Accelerate Training.
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes? Based: unclearThe paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“it can be seen that the proposed denoiser consistently outperforms CBM3D by a large margin.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It can be seen that WNNM, MLP and TNRD can outperform BM3D by about 0.3dB in PSNR.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
- It may state two separate claims as one. 13
- Probably not a substantial claim. 10
- The quote only partly states this claim. 1
Every source the paper cites for this claim was checked and supports it.
In the paper
“it has been pointed out that the combination of batch normalization and residual learning is particularly helpful for Gaussian denoising since they are beneficial to each other.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Using Batch Normalization and Residual Learning to Accelerate Training.
What the sources say
- Supportedcites [65] 65. K. Zhang, W. Zuo, Y. Chen, D. Meng, and L. Zhang. Be- yond a Gaussian denoiser: Residual learning of deep CNN for image denoising. IEEE Transactions on Image Process- ing, 2017. 4 source
Section C (pp. 4–5) says residual learning and batch normalization can benefit each other in Gaussian denoising, and reports faster convergence and better denoising performance when combined. Section B (p. 3) also describes their integration as producing fast, stable training and better denoising performance. (LLM, conf=0.95)
Strength:hedged· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It is worth emphasizing that the proposed denoiser with 24 feature maps of each layer has a comparable PSNR of 28.94dB to TNRD but delivers a faster speed.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“We empirically find using large dataset does not improve the PSNR results of BSD68 dataset [50] but can slightly improve the performance of other testing images.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
What the sources say
- Not checked — no full text of the sourcecites [50] 50. S. Roth and M. J. Black. Fields of experts. International Journal of Computer Vision, 82(2):205-229, 2009. 1, 4, 5 source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
In the paper
“It turns out that our designed model can have an average P-SNR of 29.15dB on BSD68 dataset [50], which is much better than 28.94dB of 7 layers network with traditional 3×3 convolution filter and very close to 29.20dB of 16 layers network.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 3.2. The Proposed CNN Denoiser
What the sources say
- Not checked — no full text of the sourcecites [50] 50. S. Roth and M. J. Black. Fields of experts. International Journal of Computer Vision, 82(2):205-229, 2009. 1, 4, 5 source
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As an example for the testing speed comparison, our method can super-resolve the Butterfly image in 0.5 second on GPU and 12 seconds on CPU, whereas NCSR spends 198 seconds on CPU.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.3. Single Image Super-Resolution
Strength:plain· not yet calibratedSingle caseTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: The low-resolution image can be modeled by a blurring and subsequent down-sampling operation on a high-resolution one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We can see that IDDBM3D, NCSR and MLP tend to smooth the edges and generate color artifacts.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.2. Image Deblurring
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- The authors state this as an assumption: The convolution is carried out with circular boundary conditions.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“First, although S-RCNN and VDSR achieve promising results to tackle the case with bicubic kernel, their performance deteriorates seriously when the low-resolution image are not generated by bicubic kernel (see Figure 4).”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.3. Single Image Super-Resolution
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: The low-resolution image can be modeled by a blurring and subsequent down-sampling operation on a high-resolution one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In contrast, the proposed method can recover image sharpness and naturalness.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.2. Image Deblurring
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- The authors state this as an assumption: The convolution is carried out with circular boundary conditions.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Third, both of the gray and color CNN denoiser prior based optimization methods can produce promising results.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.3. Single Image Super-Resolution
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: The low-resolution image can be modeled by a blurring and subsequent down-sampling operation on a high-resolution one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Third, CNN exploits the external prior which is complementary to the internal prior of many existing denoisers such as BM3D.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 3.1. Why Choose CNN Denoiser?
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
In the paper
“model-based optimization methods are flexible for handling different inverse problems but are usually time-consuming with sophisticated priors for the purpose of good performance”
- Also stated in: § 1. Introduction · plain
“model-based optimization methods are usually timeconsuming with sophisticated priors for the purpose of good performance [27].”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Abstract
What the sources say
- Not checked — no full text of the sourcecites [27] 27. Q. Gao and S. Roth. How well do filter-based MRFs model natural images? In Joint DAGM (German Association for Pattern Recognition) and OAGM Symposium, pages 62-72, 2012. 2 source
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Also stated in: § 1. Introduction · plain
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Second, CNN exhibits powerful prior modeling capacity with deep architecture.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 3.1. Why Choose CNN Denoiser?
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“First, the inference of CNN is very efficient due to the parallel computation ability of GPU.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 3.1. Why Choose CNN Denoiser?
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Especially noteworthy is the number of the denoisers which is much less than that of learning different models for different degradations.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Using Batch Normalization and Residual Learning to Accelerate Training.
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“discriminative learning methods have fast testing speed but their application range is greatly restricted by the specialized task.”
- Also stated in: § 1. Introduction · plain
“discriminative learning methods are usually restricted by specialized tasks.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Abstract
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Also stated in: § 1. Introduction · plain
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Such a good compromise between speed and performance over TNRD is properly attributed to the following three reasons.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one. The quote only partly states this claim.
In the paper
“and it is preferable to jointly handle RGB channels.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 3.1. Why Choose CNN Denoiser?
What the sources say
- Not matchedcites [42] 42. T. Miyata. Inter-channel relation based vectorial total varia- tion for color image recovery. In IEEE International Confer- ence on Image Processing,, pages 2251-2255, 2015. 3
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
- Not matched
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“To make a further comparison with TNRD under similar PSNR performance, we additionally provide the run time of the proposed denoiser where each middle layer has 24 feature maps.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.1. Image Denoising
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the denoiser here just acts its own role regardless of the noise type and noise level of the image to be denoised.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > Using Batch Normalization and Residual Learning to Accelerate Training.
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“To show the advantage of our design over the above two cases, we have trained three different models on noise level 25 with same training settings.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 3.2. The Proposed CNN Denoiser
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“It is worthy noting that the iterative regularization step of methods such as NCSR and WNNM actually corresponds to solving Eqn.(6a).”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.3. Single Image Super-Resolution
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“The number of iterations is set to 30 as we find it is large enough to obtain a satisfying performance.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 4.2. Image Deblurring
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
- The authors state this as an assumption: The convolution is carried out with circular boundary conditions.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“On the other hand, different from discriminative learning methods which are specialized for certain image restoration tasks, the proposed deep CNN denoiser prior based optimization method is flexible in handling various tasks while can produce very favorable results.”
§ Learning Deep CNN Denoiser Prior for Image Restoration > 5. Conclusion
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
The paper relies on this without support.
The study of integration with fast discriminative denoiser prior is still lacking.
“However, the study of integration with fast discriminative denoiser prior is still lacking.”
“However, to the best of our knowledge, the study of integration with discriminative denoiser is still lacking.”
The authors state this as an assumption.
The convolution is carried out with circular boundary conditions.
“In addition, we assume the convolution is carried out with circular boundary conditions.”
The paper relies on this without support.
The low-resolution image can be modeled by a blurring and subsequent down-sampling operation on a high-resolution one.
“In general, the low-resolution (LR) image can be modeled by a blurring and subsequent down-sampling operation on a high-resolution one.”
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 1×
“Thus, using 3×3 filter with a large depth is popularized in existing CNN network design [30, 35, 56].”
MethodVery Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 1×
“However, increasing the filter size would not only introduce more parameters but also increase the computational burden [53].”
BackgroundImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 1×
“Particularly, several CNN techniques, including Rectifier Linear Units (ReLU) [37], batch normalization [32], Adam [36], dilated convolution [63] are adopted into the network design or training.”
MethodAdam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 5×
“Particularly, several CNN techniques, including Rectifier Linear Units (ReLU) [37], batch normalization [32], Adam [36], dilated convolution [63] are adopted into the network design or training.”
MethodInfluentialImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · cited 1×
“Hence, instead of training on a small dataset consisting of 400 Berkeley segmentation dataset (BSD) images of size 180×180 [13], we collect a large dataset which includes 400 BSD images, 400 selected images from validation set of ImageNet database [20] and 4,744 images of Waterloo Exploration Database [40].”
MethodGoing deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562 · cited 1×
“Thus, using 3×3 filter with a large depth is popularized in existing CNN network design [30, 35, 56].”
MethodBatch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818 · cited 2×
“Particularly, several CNN techniques, including Rectifier Linear Units (ReLU) [37], batch normalization [32], Adam [36], dilated convolution [63] are adopted into the network design or training.”
BackgroundMethodBoyd, S., Parikh, N., Chu, E., Peleato, B., & Eckstein, J. (2011) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000016 · Cited by 17,075
Boyd, S. (2010) · now publishers, Inc. eBooks · doi:10.1561/9781601984616 · Cited by 12,917
A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
Beck, A., & Teboulle, M. (2009) · SIAM Journal on Imaging Sciences · doi:10.1137/080716542 · Cited by 11,952 · cited 1×
“Note that this property can be also employed in other optimization methods ( e.g ., iterative shrink-age/thresholding algorithms ISTA [4, 14] and FISTA [3]) as long as there involves a denoising subproblem.”
Flagged first, then by citations
6 of the 81 references couldn’t be matched to a known paper.
Research works printed but not linked: 2
- foi, 2006, Pointwise shape adaptive DCT denoising with structure preservation in luminance-chrominance space, International Workshop on Video Processing and Quality Metrics for Consumer Electronics
- danielyan, 2010, Image deblurring by augmented lagrangian with BM3D frame prior, Workshop on Information Theoretic Methods in Science and Engineering, 16
Cited by
Sorted by citations
Items 1,701–1,800 of 2,094
Face quality analysis of single-image super-resolution based on SIFT
Hu, X. G., Sun, J., Mai, Z., et al. (2019) · Signal Image and Video Processing · doi:10.1007/s11760-019-01614-1 · Cited by 1
“IRCNN [12] × ✓ ✓ Set5, Set14 ✓ × × × ✓ ×”
A Novel Attention Enhanced Dense Network for Image Super-Resolution
Niu, Z.-H., Zhou, Y., Yang, Y.-B., et al. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-37731-1_46 · Cited by 1
A Study on Digital Image Restoration Filters
Shaila Banu, S. K., Sivaparvathi, B., Munwar Ali, S. K., et al. (2020) · International Journal of Scientific Research in Science and Technology · doi:10.32628/ijsrst120724 · Cited by 1
Image Restoration Via Data-Dependent Proximal Averaged Optimization
Mu, P., Chen, J., Liu, R., Zhong, W., Fan, X., & Luo, Z. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9053547 · Cited by 1
“In general, the existing predominant IR methods can be roughly classified into two categories: designed prior-based methods [6, 7, 8] and learning-based methods [3, 9, 10, 11, 12].”
MethodInfluentialMulti-frame super resolution via deep plug-and-play CNN regularization
Zhao, S., & Hu, L. (2020) · Journal of Inverse and Ill-Posed Problems · doi:10.1515/jiip-2019-0054 · Cited by 1
Deconvolution of Image Sequences with a Learning FFT-based Approach
Koshelev, I., Somov, A., Lefkimmiatis, S., & Rodríguez-Sánchez, A. J. (2020) · IEEE International Symposium on Industrial Electronics (ISIE) · doi:10.1109/isie45063.2020.9152463 · Cited by 1
“…towards both improving the restora-978-1-7281-5635-4 tion quality and increasing the computational efficiency comes from the combination of classical deconvolution methods in the frequency domain and the recent advances in the field of machine learning and especially deep learning [16]–[19].”
Deep multi‐level up‐projection network for single image super‐resolution
Shen, Y., Zhang, L., Chen, Y., et al. (2020) · IET Image Processing · doi:10.1049/ipr2.12014 · Cited by 1
“Therefore, IRCNN [3] introduced dilated convolution to tackle low-level vision tasks, thus enlarging the receptive field of a network without the sacrifice of computational cost.”
MethodSelf-citationDeep Learning and RFID System Physical Anti-Collision
Yu, X., Zhao, Z., & Zhang, X. (2021) · Springer eBooks · doi:10.1007/978-981-16-0835-3_6 · Cited by 1
REMOVAL OF MOIRÉ PATTERNS USING CONVOLUTIONAL NEURAL NETWORKS AND PARTICLE SWARM OPTIMIZATION
Mounika, G., & Rajashekar, K. (2021) · Journal of Emerging Technologies and Innovative Research · Cited by 1
Residual Fractal Network for Single Image Super Resolution by Widening and Deepening
Gu, J., Qu, Z., Wang, X., et al. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9412184 · Cited by 1
“As in [6], [22], we also compare 3× SR results with 9 stateof-the-art methods with blur-down (BD) degradation model: SPMSR [39], SRCNN [1], FSRCNN [37], VDSR [2], IRCNN [40], SRMDNF [29], RDN [9], RCAN [6], SAN [22].”
ResultHierarchically Aggregated Residual Transformation for Single Image Super Resolution
Hou, Z., & Kung, S. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9411955 · Cited by 1
“Following the same setup in [3], [18], we compare HARTnet with SRCNN [1], VDSR [6], IRCNN [31], SRMD [32], RDN [3], SRFBN [18] under the challenging DN degradation (x3 SR, σ = 30 denoising).”
MethodMohammadi, N., Doyley, M. M., & Çetin, M. (2021) · doi:10.1109/ssp49050.2021.9513799 · Cited by 1
“This group of methods including Plug-andPlay (PnP) [17], [18] and regularization by denoising (RED) [19] learn a data-driven denoiser and then plug it as the regularizer proximal operator or regularizer gradient into the constrained optimization task .”
MethodRobust Image Denoising with Texture-Aware Neural Network
Fu, B., Wang, L., & Luo, Z. (2021) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme51207.2021.9428158 · Cited by 1
“With the continuously development of deep learning technology, its strong self-learning ability is applied in high-level computer vision [13, 14, 15, 16] and low-level computer vision tasks such as image restoration and image segmentation[17, 18].”
BackgroundMethodEffect of image degradation on performance of Convolutional Neural Networks
Aljarrah, I. A. (2021) · International Journal of Communication Networks and Information Security (IJCNIS) · doi:10.54039/ijcnis.v13i2.4946 · Cited by 1
Plug-and-Play ADMM Using Deep Image Prior for Image Denoising
Xie, Z., & Liu, L. (2021) · International Conference on Pattern Recognition and Artificial Intelligence (PRAI) · doi:10.1109/prai53619.2021.9551100 · Cited by 1
Image Denoising with Self-adaptive Multi-UNET Valve
Thesia, Y., Suthar, M., Pandya, T., et al. (2021) · Advances in intelligent systems and computing · doi:10.1007/978-981-16-2709-5_49 · Cited by 1
Multi-level Residual Attention Network for Speckle Suppression
Lei, Y., Liu, S., Zhang, L., et al. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88013-2_24 · Cited by 1
Self-citationA Hybrid Sparse Representation Model for Image Restoration
Zhou, C., Kong, Y., Zhang, C., et al. (2022) · Sensors · doi:10.3390/s22020537 · Cited by 1
“In order to comprehensively evaluate the performance of the proposed model for image restoration, the proposed HSR model was also compared with algorithms based on deep learning [42,43].”
MethodResultInfluentialCNN-Based Deblurring of THz Time-Domain Images
Ljubenović, M., Bazrafkan, S., Paramonov, P., Beenhouwer, J. D., & Sijbers, J. (2022) · Communications in computer and information science · doi:10.1007/978-3-030-94893-1_22 · Cited by 1
Jeong, J., Jang, H., Caesary, D., Joung, I. S., Cho, A., Yoon, D., & Nam, M. J. (2022) · Journal of the Korean Society of Mineral and Energy Resources Engineers · doi:10.32390/ksmer.2022.59.4.379 · Cited by 1
“…분포(posteriori distribution)로서의 해인 는 일 반적으로 최대 사후 확률(maximum a posteriori, MAP) 문 제를 해결함으로써 다음과 같이 계산할 수 있다(Zhang et al., 2017b; Kataoka and Yasuda, 2019; Delon and Houdard, 2018). arg max log log (2) 여기서 log 는 관측 값의 로그 가능성, log 는 선행 값을 나타낼 수 있으며, (2)식은 다음과 같이 최소화 하여 에너지 함수로 재구성할 수 있다(Zhang et…”
Reinforcement Learning Based Plug-and-Play Method for Hyperspectral Image Reconstruction
Fu, Y., & Zhang, Y. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20497-5_38 · Cited by 2
Li, N., & Liu, C. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.09878 · Cited by 1
“Our model MC-ISTA-Net is compared with the state of art SR method: RCAN + [28], IRCNN[35], CS method:.”
MethodHigh-Quality Automatic Foreground Extraction Using Consensus EquilibriumPreprint
Wang, X., Juang, J., & Chan, S. H. (2018) · arXiv (Cornell University) · Cited by 1
Deep Likelihood Network for Image Restoration with Multiple Degradation LevelsPreprint
Guo, Y., Lu, M., Zuo, W., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.09105 · Cited by 1
“It also lies in the category of MAP inference guided discriminative learning [62].”
BackgroundSelf-citationPerformance Analysis of Plug-and-Play ADMM: A Graph Signal Processing PerspectivePreprint
Chan, S. H. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.00020 · Cited by 1
Online Regularization by Denoising with Applications to Phase RetrievalPreprint
Wu, Z., Sun, Y., Liu, J., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.02040 · Cited by 1
“Recent work has demonstrated the benefit of using denoisers as priors for solving imaging inverse problems [8, 12, 23, 26, 27, 37, 40, 41, 43, 49].”
BackgroundConverged Deep Framework Assembling Principled Modules for CS-MRIPreprint
Liu, R., Zhang, Y., Cheng, S., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1910.13046 · Cited by 1
“We regard the artifacts introduced by the fidelity mapping F as unknown noise, and thus employ a residual denoising network with shortcuts (IRCNN) [36] as the data-driven module of our framework:”
MethodInfluentialRevisiting Non Local Sparse Models for Image RestorationPreprint
Lecouat, B., Ponce, J., & Mairal, J. (2019) · HAL (Le Centre pour la Communication Scientifique Directe) · Cited by 1
Gated Texture CNN for Efficient and Configurable Image DenoisingPreprint
Imai, K., & Miyata, T. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.07042 · Cited by 1
“In particular, GTCNN-D6 achieves new state-of-the-art denoising performance but is 4.8 times smaller than MWCNN [22] (see Table 1 for more details) 46, 29, 28, 17, 18, 2, 49, 22, 7, 14].”
BackgroundInfluentialAttention Prior for Real Image Restoration.Preprint
Anwar, S., Barnes, N. M., & Petersson, L. (2020) · arXiv (Cornell University) · Cited by 1
Zhao, Y., Ossowski, J., Wang, X., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.05149 · Cited by 1
“Integration of the two approaches has delivered promising results in tasks such as object recognition [18], [20], image captioning [21], [22] and image restoration [23], [24].”
BackgroundBlind Image Restoration with Flow Based PriorsPreprint
Helminger, L., Bernasconi, M., Djelouah, A., Groß, M. H., & Schroers, C. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2009.04583 · Cited by 4
“…deep learning approaches for dedicated image restoration problems such as super-resolution (Wang et al., 2018; Zhang et al., 2018), denoising (Zhang et al., 2017a), inpainting (Pathak et al., 2016) or a combination of them (Park & Mu Lee, 2017), one important drawback is the need for…”
BackgroundLearning Deep Interleaved Networks with Asymmetric Co-Attention for Image RestorationPreprint
Li, F., Cong, R., Bai, H., He, Y., Zhao, Y., & Zhu, C. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.15689 · Cited by 1
“Our DIN are compared with VDSR [10], IRCNN [14], EDSR [11], RDN [24], and SRFBN [37].”
MethodDeep Denoising of Flash and No-Flash Pairs for Photography in Low-Light EnvironmentsPreprint
Xia, Z., Gharbi, M., Perazzi, F., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.05116 · Cited by 1
“[3] were the first to show the ability of even shallow multi-layer perceptrons to to outperform traditional methods such as BM3D [7], and more recent approaches utilizing deeper networks and complex architectures [24, 34, 36, 39, 40, 41] have since led to further improvements in reconstruction accuracy.”
BackgroundTowards Boosting the Channel Attention in Real Image Denoising : Sub-band Pyramid AttentionPreprint
Li, H., Wu, H., Chen, X., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.12481 · Cited by 1
“Previous CNN-based methods [1], [2], [6] outperform the traditional methods [7], [8] due to the powerful learning ability of CNNs.”
MethodDigital Image Restoration in Matlab: A Case Study on Inverse and Wiener FilteringPreprint
Khan, M. M. R., Sakib, S., Arif, R. B., et al. (2018) · Preprints.org · doi:10.20944/preprints201811.0566.v1 · Cited by 1
Learning local regularization for variational image restorationPreprint
Prost, J., Houdard, A., Almansa, A., et al. (2021) · arXiv (Cornell University) · Cited by 1
Machine Learning aided k-t SENSE for fast reconstruction of highly accelerated PCMR dataPreprint
Kowalik, G. T., Montalt‐Tordera, J., Steeden, J. A., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.07541 · Cited by 1
“For example, aliasing artefacts can be suppressed through ML based denoising of MR images [9], as long as these artefacts are incoherent in nature.”
BackgroundMethodDWDN: Deep Wiener Deconvolution Network for Non-Blind Image DeblurringPreprint
Dong, J., Roth, S., & Schiele, B. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.09962 · Cited by 1
“The images deblurred by the methods [3], [6], [15], [17], [18], [45], [50], [57] contain severe artifacts as shown in (b) and (d)–(j) .”
BackgroundMethodInfluentialConditional Meta-Network for Blind Super-Resolution with Multiple Degradations.Preprint
Yin, G., Wang, W., Yuan, Z., et al. (2021) · arXiv (Cornell University) · Cited by 1
Feedback Pyramid Attention Networks for Single Image Super-ResolutionPreprint
Wu, H., Gui, J., Zhang, J., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.06966 · Cited by 1
“We compare our FPAN with some state-of-the-art methods: SRCNN [4], FSRCNN [51], VDSR [5], ”IRCNN G“ [52], ”IRCNN C“ [52], RDN [17], SRFBN [30].”
MethodMoDIR: Motion-Compensated Training for Deep Image Reconstruction without Ground TruthPreprint
Gan, W., Sun, Y., Eldeniz, C., Liu, J., An, H., & Kamilov, U. S. (2021) · arXiv (Cornell University) · Cited by 1
DRFENet: Toward a residual network of dilated convolution for image denoisingPreprint
Zhong, R., Zuo, M., Zhang, Q., & Yan, W. (2022) · Research Square · doi:10.21203/rs.3.rs-1478737/v1 · Cited by 1
Convergent plug-and-play with proximal denoiser and unconstrained regularization parameter.Preprint
Hurault, S., Chambolle, A., Leclaire, A., et al. (2023) · Research Square · doi:10.21203/rs.3.rs-3462535/v1 · Cited by 1
Zhang, Q., He, L., Wang, Y., Deng, L.-J., & Liu, J. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4619731
ReTiDe: Real-Time Denoising for Energy-Efficient Motion Picture Processing with FPGAs
Li, C. L., Bled, C., Fernandez, R., et al. (2025) · doi:10.1145/3756863.3769710 · Cited by 1
“Subsequent models built on this foundation, such as IRCNN [Zhang et al. 2017b] with dilated convolutions and FFDNet [Zhang et al. 2018], which incorporates user-provided noise levels and subsampled inputs.”
Multi Task Denoiser Training for Solving Linear Inverse Problems
Bled, C., & Pitié, F. (2025) · doi:10.1145/3756863.3769703 · Cited by 1
Neural networks regularization through representation learning
Belharbi, S. (2018) · theses.fr (ABES) · doi:10.70675/0e1f98e2zabc7z497bzaf40z433d33a78b60 · Cited by 3
“Nowadays, different deep neural models including feedforward and recurrent networks have been successfully applied to different tasks including: computer vision [88, 244], speech recognition [106, 380], natural language processing and machine translation [400, 387], visual art processing8 [420, 395], recommendation sy…”
ResultDebarnot, V. (2020) · theses.fr (ABES) · doi:10.70675/8ef90aedz4ba9z4d2azbd83z416307995a7a · Cited by 1
Fall detection and activity recognition using stereo low-resolution thermal imaging
Zoetgnande, Y. W. K. (2020) · theses.fr (ABES) · doi:10.70675/f71894b5z8fa1z4817z9cf8z7c1abada8e37 · Cited by 1
Hybrid Pixel-Unshuffled Network for Lightweight Image Super-resolution
Sun, B., Zhang, Y., Jiang, S., & Fu, Y. (2023) · AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) · doi:10.1609/aaai.v37i2.25333 · Cited by 94
Efficient $\ell ^0$ Gradient-Based Super-Resolution for Simplified Image Segmentation
Cascarano, P., Calatroni, L., & Piccolomini, E. L. (2021) · IEEE Transactions on Computational Imaging · doi:10.1109/tci.2021.3070720
“In particular, in [23] the authors present a Plug-and-Play (PnP) framework [24] which exploits deep convolutional neural network denoisers embedded in standard optimisation algorithms, such as Alternating Direction of Multipliers (ADMM) or Half-Quadratic Splitting (HQS).”
BackgroundMethodA Deep Convolution Neural Network Based Model for Enhancing Text Video Frames for Detection
Sunil, C., Chethan, H. K., Raghunandan, K. S., et al. (2018) · Advances in intelligent systems and computing · doi:10.1007/978-3-319-76348-4_42
Simultaneous sparse coding with Laplacian scale mixture prior for image restoration
Lu, X., Lü, X., & Zuo, Y. (2018) · Journal of Electronic Imaging · doi:10.1117/1.jei.27.5.053009
Prior-guided deep residual propagation for non-blind image deconvolution
Ma, L., Jiang, Z., Liu, R., & Fan, X. (2018) · International Conference on Internet Multimedia Computing and Service · doi:10.1145/3240876.3240896
“[18] embedded a learning-based CNN denoiser into the step of half-quadratic optimization algorithms.”
BackgroundMethodPixel Convolutional Networks for Skeleton-Based Human Action Recognition
Chang, Z., Wang, J., & Han, L. (2018) · Communications in computer and information science · doi:10.1007/978-981-13-2853-4_40
Single-image reflection removal : from computational imaging to deep learning
Wan, R. (2018) · doi:10.32657/10220/47556
“Prior works that use deep learning to solve the inverse imaging problems [92, 93] or layer separation problems [94] mainly optimize the whole network on patches with resolution n×n cropped from the whole images.”
BackgroundA Textural Layer L0 Gradient Minimization based De-noise Method for Scene Text Images
Wang, S., Li, X., Dong, N., et al. (2018) · International Conference on Signal Processing · doi:10.1109/icsp.2018.8652439
“Zhangkai and others [5, 6] proposed an image denoising method based on residual learning, which combines deep learning framework with normalization method.”
MethodLearning deep neural networks in blind deblurring framework.Preprint
Wu, J., Di, X., Huang, J., & Zhang, Y. (2019) · arXiv (Cornell University)
Novel Image and Video Super-Resolution Relying on Denoising Algorithms
Brifman, A., & Elad, M. (2018)
Super-Resolution Reconstruction Using Deep Learning: Should We Go Deeper?
Kostrzewa, D., Piechaczek, S., Hrynczenko, K., et al. (2019) · Communications in computer and information science · doi:10.1007/978-3-030-19093-4_16
Deep CNN jointing low-high level feature for image super-resolution
Song, X., Liu, W., Liu, J., et al. (2019) · International Conference on Graphics and Image Processing (ICGIP) · doi:10.1117/12.2524412
Learning deep optimizer for blind image deconvolution
Luo, Z., Chen, S., & Qian, Y. (2019) · International Journal of Wavelets Multiresolution and Information Processing · doi:10.1142/s0219691319500449
Application of a Scaled MNIST Dataset Blended with Natural Scene Background on ResNet
Marinov, A., Mtetwa, N., & Larijani, H. (2019) · International Conference on Big Data and Education (ICBDE) · doi:10.1145/3322134.3322147
“- image restoration [5] – the process allows an image restoration from a low-level resolution to provide better quality of the image.”
BackgroundIntersections and sums of sets for the regularization of inverse problems
Peters, B. (2019) · cIRcle (University of British Columbia) · doi:10.14288/1.0378709
An Adaptive Pointwise P-norm Regularization for Image Deconvolution
Ma, B., Hu, X., & Peng, S. (2019) · IEEE International Conference on Digital Signal Processing (DSP) · doi:10.1145/3316551.3316563
Optimal Combination of Image DenoisersPreprint
Choi, J. H., Elgendy, O. A., & Chan, S. H. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1711.06712 · Cited by 1
“Conventional approach is to either truncate the noise level to the nearest trained level [23] or to train the network with a large number of examples covering all noise levels [22].”
Compressed Sensing MRI via a Multi-scale Dilated Residual Convolution NetworkPreprint
Dai, Y., & Zhuang, P. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1906.05251
Li, Y., Gui, G., & Cheng, X. (2019) · arXiv (Cornell University)
Comparing and Combining MLPs and CNNs for Image Denoising
Zhang, D., Zhao, L., Xu, D., et al. (2018) · IEEE International Conference on Computer and Communications (ICCC) · doi:10.1109/compcomm.2018.8780788
“[12], [13], [14] are valid Convolutional Neural Network (CNNs) methods for image denoising.”
BackgroundAn inquiry on image enhancement methods of remote sensing images
Liu, J., Pu, X., Wen, D., et al. (2019) · International Conference on Digital Image Processing · doi:10.1117/12.2540274
On the Convergence of ADMM with Task Adaption and Beyond.Preprint
Liu, R., Mu, P., & Zhang, J. (2019) · arXiv (Cornell University)
13.1: Invited Paper: Multigrid Backprojection Super‐Resolution and Deep Filter Visualization
Chen, G., Zhu, D., Zhang, L., et al. (2019) · SID Symposium Digest of Technical Papers · doi:10.1002/sdtp.13409
Self-citationA Network-Driven Prior Induced Bregman Model for Parallel MR Imaging
Li, G., Liu, Y., Zhang, M., et al. (2019) · doi:10.1109/embc.2019.8856914
“Especially, rather than focused on the end-to-end networks, several researchers paid attention to embed the deep learning networks into the model-based optimization for the IR problems [19]-[21].”
BackgroundCross-Scale Residual Network for Multiple Tasks:Image Super-resolution, Denoising, and DeblockingPreprint
Zhou, Y., Du, X., Zhang, Y., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1911.01257
“[36],PGPD15 [53] DnCNN16 [12], IRCNN17 [14], RED18 [51], MemNet17 [13] and FOCNet19 [54].”
MethodInfluentialDense Activation Network for Image Denoising
Shen, Y., Zhang, L., Lou, S., et al. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-31723-2_7
Self-citationDeep CNN for Single Image Super Resolution Using Skip Connections
Basavaraju, S., Bahuguna, S., Gaj, S., et al. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-34869-4_24
A Comparative Study of FFDNet and Curvelet Thresholding for Image Denoising
Srinivas, N., Sai, N., & Akkala, P. (2019)
Deep Learning for 3D Classification Based on Point Cloud with Local Structure
Song, Y., Li, X., & Gao, L. (2019) · IEEE International Conference on Information Communication and Signal Processing (ICICSP) · doi:10.1109/icicsp48821.2019.8958558
“Although deep learning has achieved remarkable results on 2D images [3, 4], few deep learning algorithms use point cloud as the input because of its irregular format.”
BackgroundDnM3Net: Multi-Scale & Multi-Level Shuffle-CNN Via Multi-Level Attention for Image Denoising
Cao, Y., He, J., Zhang, Y., et al. (2019) · doi:10.1109/iucc/dsci/smartcns.2019.00075
“Dataset σ BM3D [30] TNRD [31] DnCNN [8] IRCNN [32] RED [9] MemNet [33] MWCNN [12] DnM3Net”
BackgroundMethodInfluentialRethink Gaussian Denoising Prior for Real-World Image Denoising
Wang, T., Huan, J., Li, B., & Hu, K. (2019) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai.2019.00244
“Most of existing related methods are designed for Gaussian denoising [1]–[8].”
MethodBoosting the Performance of Plug-and-Play Priors via Denoiser ScalingPreprint
Xu, X., Liu, J., Sun, Y., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.11546
“Recent results have shown that by using advanced image denoisers in iterative image reconstruction, PnP algorithms achieve state-of-the-art performance in many imaging problems [17–23].”
BackgroundBeyond Camera Motion Removing: How to Handle Outliers in DeblurringPreprint
Chang, M., Yang, C., Feng, H., et al. (2020) · arXiv (Cornell University)
Weak Texture Information Map Guided Image Super-resolution with Deep Residual NetworksPreprint
Fu, B., Wang, L., Wu, Y., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.00451
“Among these categories of methods, the deep neural network based methods[15,16,17,18,19,20,21,22] have achieved significant improvements.”
MethodA Partially Pre-Trained Dense Network for Image Super-Resolution
Zhang, Y., Liu, P., Chen, Q., et al. (2019) · International Conference on Systems and Informatics (ICSAI) · doi:10.1109/icsai48974.2019.9010137
“SRResNet [24], VDSR [25] and IRCNN [26] propose residual learning SR networks.”
BackgroundWhen Autonomous Systems Meet Accuracy and Transferability through AI: A SurveyPreprint
Zhang, C., Wang, J., Yen, G. G., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.12948
“In addition, image restoration often uses methods, like image sharpening, in which the input image and output image remain the same size, although the output quality can be improved [181].”
MethodUnsupervised Deep Learning for MR Angiography with Flexible Temporal ResolutionPreprint
Cha, E., Chung, H., Kim, E. Y., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.13096
“This implies that in contrast to the inverse formulation using deep learning prior [27], [28] that incorporate the physics-driven information during the run-time reconstruction and network architectures, our cycleGAN formulation shows that the physics-driven constraint can be incorporated in…”
ResultLearning regularization and intensity-gradient-based fidelity for single image super resolutionPreprint
Hu, L., & Zhao, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.10689
“Recent study [38-40] shows that the regularizer can be replaced by the implicit prior learned from the neural network.”
BackgroundMethodInfluentialDeep Learning Based Single Image Super ResolutionPreprint
Vaghela, H. (2020) · doi:10.36227/techrxiv.12059178.v1
“As I have mentioned two categories of Linear networks, Super-resolution Convolution Neural Network (SRCNN) [14] [15], VDSR [16], DnCNN [17] and IrCNN [18] are examples of early up-sampling design where they upsample the given LR image into required resolution of HR image and then learn hierarchical feature representat…”
BackgroundMethodInvestigating Image Applications Based on Spatial-Frequency Transform and Deep Learning TechniquesPreprint
Zheng, Q., Qiu, H., Memmi, G., & Isabelle Bloch, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.02756
“One of the most popular issues is to use the DL model to enhance the image quality [20] which can also be seen as to remove the noise of the image [21].”
BackgroundTime accelerated image super-resolution using shallow residual feature representative networkPreprint
Ajith, M., Kurup, A. R., & Martínez‐Ramón, M. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.04093
The Power of Triply Complementary Priors for Image Compressive SensingPreprint
Zha, Z., Yuan, X., Zhou, J. T., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.07902
“(13), which is an efficient but effective CNN-based denoisier.”
BackgroundMethodInfluentialOne Size Fits All: Can We Train One Denoiser for All Noise Levels?Preprint
Gnansambandam, A., & Chan, S. H. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.09627 · Cited by 6
“The competitiveness of these “one size fits all” denoisers compared to the best individually trained denoisers has been demonstrated in (Zhang et al., 2017; 2018; Mao et al., 2016a; Remez et al., 2017).”
BackgroundMethodInfluentialLi, Y., Gui, G., & Cheng, X. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.04699
“Image restoration (IR) problem [1][2][3] is a classic yet active topic in low-level vision.”
BackgroundIdentity Enhanced Residual Image DenoisingPreprint
Anwar, S., Huynh, C. P., & Porikli, F. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.13523
“So far, state-of-the-art denoising methods, such as DnCNN ([56]), IrCNN ([57]) and BM3D ([15]) etc.”
BackgroundMethodInfluentialDeep Spatial-angular Regularization for Compressive Light Field Reconstruction over Coded AperturesPreprint
Guo, M., Hou, J., Jin, J., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.11882
“quality constraint into the objective function as a penalty term, i.e., min x;v 1 2 kl Axk2 2 + kx vk 2 + J(v); (5) where >0 is a penalty parameter. Based on the half quadratic splitting method [5,43], the optimization problem in Eq. (5) can be solved by alternatively solving the following two sub-problems until co…”
MethodUnpaired Deep Learning for Accelerated MRI using Optimal Transport Driven CycleGANPreprint
Oh, G., Sim, B., Chung, H., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.12967
“In contrast to the recent inverse problem approaches using deep learning prior [42], [43], which incorporate physics-driven information during the run-time reconstruction, our cycleGAN formulation shows that the physics-driven constraint can be incorporated in the training phase of a feed-forward deep neural networks…”
ResultDeep Generative Modeling Based Retinal Image Analysis
Sengupta, S. (2020) · UWSpace (University of Waterloo)
Data-aware Computational Imaging Acceleration
Mahmoud, M. M. H. (2019) · TSpace (University of Toronto)
Failure Recognition for Switch Machines Based on Machine Learning
Hu, E., Zhu, C., & Li, C. (2019) · CAA Symposium on Fault Detection, Supervision and Safety for Technical Processes (SAFEPROCESS) · doi:10.1109/safeprocess45799.2019.9213416
Metadata
- Authors
- Zhang, K., Zuo, W., Gu, S., & Zhang, L.
- Affiliations
- Hong Kong Polytechnic University (HK)
- Harbin Institute of Technology (CN)
- Venue
- 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2017-07
- Pages
- 2808-2817
- OpenAlex
- W2613155248
- Cited by
- 2,116
- References
- 81
- Citations by year
- 2026: 80 · 2025: 191 · 2024: 248 · 2023: 293 · 2022: 272 · 2021: 338 · 2020: 297 · 2019: 240 · 2018: 135 · 2017: 10
- FWCI
- 49.67
- Type
- conference paper
- Language
- en
- Open access
- green
- hdl.handle.net
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- PolyU Institutional Research Archive (Hong Kong Polytechnic University) (submittedVersion)
- TUScholarShare (Temple University) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Signal Denoising Methods
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image Processing Techniques
- Physical Sciences › Engineering › Media Technology › Advanced Image Fusion Techniques
- Keywords
- image restoration, model-based optimization, discriminative learning, variable splitting, inverse problems, low-level vision, Gaussian denoising, image deblurring
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
