Pith. sign in

REVIEW 4 major objections 4 minor 39 references

Frequency-domain Learning with Kernel Prior for Blind Image Deblurring

T0 review · 4 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A frequency-domain Transformer that fuses an estimated blur-kernel prior improves blind deblurring and out-of-domain generalization.

desk verdict A new frequency-domain kernel fusion module with promising out-of-domain numbers, but an unexplained 1.66 dB gap between the ablation baseline and the retrained backbone makes the core contribution hard to evaluate. read the letter →

arxiv 2504.14664 v1 pith:2FJBGNYG submitted 2025-04-20 cs.CV

Jixiang Sun , Fei Lei , Jiawei Zhang , Wenxiu Sun , Yujiu Yang This is my paper · ORCID
classification cs.CV
keywords blindimagedeblurringkernelpriorfrequencydomainTransformerout-of-domaingeneralizationIntegrationModuledeconvolutiontheoremrestoration
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

This paper tries to establish that blind image deblurring networks generalize better when they are given an explicit estimate of the blur kernel, and that the most effective way to inject that estimate is to fuse it in the frequency domain rather than by simple channel concatenation. The authors build a kernel estimation module that predicts pixel-wise kernels from the blurry image and a Frequency Integration Module that combines those kernel features with image features through a frequency-domain attention-like operation before each block of a frequency-based Transformer backbone. They report state-of-the-art results on GoPro, HIDE, and RealBlur, with the largest gains appearing on out-of-domain data: 34.21 dB on GoPro, 32.08 on HIDE, 36.59 on RealBlur-R, and 29.01 on RealBlur-J. The importance of the claim is that kernel prior is content-independent, so it may be a practical route to making learned deblurring less dataset-bound.

What carries the argument

The Frequency Integration Module (FIM) is the load-bearing mechanism. It projects image and kernel features into query, key, and value triples and computes Frequency Attention as $FA(Q,K,V)=L(F^{-1}(F(Q)\odot F(K)))\odot V$, where $F$ is the Fourier transform, $F^{-1}$ its inverse, $L$ layer normalization, and $\odot$ element-wise multiplication. This mimics the deconvolution theorem, that blur in the spatial domain is pointwise multiplication in the frequency domain, and lets the kernel act as a frequency-domain mask on the image feature. The module is inserted before each encoder and decoder block of a frequency-domain Transformer backbone, at multiple kernel-feature scales, with a residual connection that preserves the original feature; a U-Net-style kernel estimator trained by reblurring the sharp image with the predicted kernel provides the pixel-wise kernel prior.

What would settle it

Retrain the FFTFormer backbone under the paper's own stage-II and stage-III schedules without any kernel prior, measure GoPro PSNR, and then add FIM with the same schedule; if the FIM-augmented model does not beat that matched baseline by roughly the claimed margin, the kernel-prior effect is not established. A simpler check is to explain the 1.66 dB gap between the 34.11 dB retrained backbone and the 32.45 dB ablation baseline.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that the blur kernel, estimated as a spatially varying map and converted to a frequency-domain conditioning signal, supplies information about the degradation that is independent of image content. When fused by Frequency Attention, the Hadamard product of Fourier-transformed image query and kernel key, followed by inverse transform and element-wise modulation of the value, the kernel prior makes a frequency-domain Transformer both sharper on in-domain data and substantially more robust on out-of-domain benchmarks. The authors show the fusion works best in both encoder and decoder, at multiple scales, with a residual connection inside the module, and that giving the module ground-truth kernels improves results further, indicating the module itself is sound and the bottleneck is kernel estimation accuracy.

Load-bearing premise

The load-bearing assumption is that the 32.45 dB no-kernel baseline in the ablation study is the same backbone trained under the same conditions as the FIM-augmented model; if that baseline is weaker than the backbone's true retrained performance, the reported gains from the kernel prior are partly an artifact of the comparison.

Editorial extensions

If this is right

  • On the GoPro-trained model, adding the kernel prior raises PSNR from 34.11 to 34.21 dB on GoPro, from 31.62 to 32.08 dB on HIDE, and from 35.46 to 36.59 dB on RealBlur-R, with SSIM at least matching on GoPro and improving on HIDE, RealBlur-R, and RealBlur-J.
  • Out-of-domain generalization improves more than in-domain performance, which is the pattern the authors predict from the kernel prior being content-independent.
  • Ablations show that both the residual connection and multi-scale encoder-decoder insertion are required; removing the residual connection turns FIM from helpful into harmful.
  • When true kernels replace estimated ones, PSNR rises further on all four synthetic test sets, so the framework can absorb better kernel estimators without architectural change.
  • The full model adds moderate cost over the backbone: 24.4 million parameters and 250.2 GFLOPs versus 16.6 million and 131.8 GFLOPs for the retrained FFTFormer.

Reading between the lines

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

  • If the mechanism is as content-independent as claimed, a similar frequency-domain kernel-fusion block should transfer to other restoration tasks that have an underlying degradation kernel, such as super-resolution or atmospheric turbulence correction; this is a direct testable extension the paper does not run.
  • The pattern of gains, largest on RealBlur-R and smaller on GoPro, suggests the practical value of kernel priors may lie less in pushing the in-domain ceiling and more in stabilizing deployment across camera systems; a controlled cross-dataset sweep with multiple training sets could quantify this.
  • One internal comparison needs scrutiny before attributing the gains: Table 3's no-kernel baseline is reported at 32.45 dB, while the same backbone retrained by the authors is stated at 34.11 dB; whether the ablation baseline is the identical network under identical training determines how much of the improvement is actually due to the kernel prior.
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

4 major / 4 minor

Summary. The paper proposes a new architecture for blind image deblurring that combines an explicit blur kernel estimation module with a frequency-domain Transformer backbone. The kernel prior is fused into the network through a Frequency Integration Module (FIM) that performs a cross-attention-like operation in the Fourier domain. Training is conducted in three stages: pretraining the kernel estimator with a reblurring loss, training the backbone and FIM with the kernel estimator frozen, and joint fine-tuning with an additional reblurring loss. The authors evaluate on GoPro, HIDE, RealBlur-R, and RealBlur-J, reporting state-of-the-art results and claiming improved out-of-domain generalization. The central claim is that the explicit kernel prior improves deblurring performance and generalization.

Significance. If the reported results are reproducible, the method provides a modest improvement over the FFTFormer backbone on the in-domain GoPro benchmark (+0.10 dB) and larger gains on out-of-domain HIDE (+0.46 dB) and RealBlur-R (+1.13 dB). The idea of using a learned kernel prior for domain generalization is of interest to the image restoration community. The paper does not provide code or pre-trained models, and no machine-checked proofs or parameter-free derivations are involved. The main significance hinges on the ablation study in Table 3, which currently contains an unexplained baseline discrepancy that must be resolved before the improvement can be attributed to the kernel prior.

major comments (4)
  1. [Table 3, row 1 vs Section 4.2.1] The first row of Table 3 is described as 'end-to-end training without introducing kernel prior' and reports 32.45 dB PSNR on GoPro, yet Section 4.2.1 states that the authors retrained the same FFTFormer backbone and obtained 34.11 dB PSNR. This 1.66 dB discrepancy is not explained. The Table 3 caption says all options are trained with the same setting, but the relationship between that setting and the Section 4.2.1 retraining protocol is not specified. If the baseline is not the backbone under the full model's training protocol, the 0.74 dB improvement attributed to FIM (row 6, 33.19 dB) is not interpretable. The value 32.45 dB also exactly matches MIMO-UNet+'s GoPro PSNR in Table 1, raising the possibility of a transcription error. Please clarify the training protocol and correct or explain the numbers.
  2. [Section 3.2, Eq. (3)] The proposed Frequency Attention operation is not fully specified. F(Q) and F(K) are complex-valued, and their pointwise product is complex; the inverse FFT generally produces a complex output, but layer normalization and element-wise multiplication with V are typically defined for real tensors. The paper does not state whether the real part is taken, nor how the kernel feature B_in is projected to match the spatial and channel dimensions of the image feature Q. These implementation details are essential for reproducibility and for verifying that the operation behaves as intended.
  3. [Section 4.3, Table 4] The non-blind ablation compares the proposed method with ground-truth kernels against a 'Baseline' (without kernel prior) on synthetic datasets. However, the synthetic dataset construction (extracting kernels from the original datasets and reblurring the corresponding sharp images) is described only briefly. The reported PSNR values for the same dataset names (e.g., GoPro Syn. 28.48 dB vs. GoPro test 32.45 dB in Table 3) are very different, and it is unclear whether these synthetic sets are derived from the training or test partitions. Without this clarification, the non-blind result does not provide a clean measure of the potential benefit of a more accurate kernel estimator.
  4. [Section 3.3, Eq. (7)] In stage III, the reblurring loss L_KE is computed using the predicted deblurred image x_hat_n from the backbone rather than the ground truth x_n. Since x_hat_n is itself an output being trained, the kernel estimator is optimized against a target that depends on the current deblurring quality. The paper does not analyze the potential interaction or instability of this joint optimization, and no sensitivity analysis is provided for the weights lambda_1 and lambda_2 (both set to 0.1). This is a load-bearing point because the reblurring loss is central to the self-supervised kernel estimation that supports the generalization claim.
minor comments (4)
  1. [Section 4.2.1] The heading 'Qualitative Comparison' is misleading because the section reports quantitative results (PSNR/SSIM); the subsequent Section 4.2.2 'Visual Comparison' is the qualitative part. Please rename the sections accordingly.
  2. [Table 3] The use of double quotes to indicate which options are enabled is unconventional and hard to read. Consider using checkmarks or a different formatting convention.
  3. [Section 4.3] The naming of synthetic datasets is inconsistent: the text says 'synthetic datasets are respectively based on GoPro, HIDE, RealBlur-R and RealBlur-J', while Table 4 uses 'GoPro (Syn.)', 'HIDE (Syn.)', 'RB-R (Syn.)', and 'RB-J (Syn.)'. Please standardize the terminology and clarify the relationship to the original test sets.
  4. [Section 4.1] The paper states that stage II 'follows the training settings of [12]' but does not report the number of training iterations for stage II. Since the three-stage training schedule is a key part of the method, this information should be included.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the kernel-prior pipeline is an empirical architecture with independent benchmark comparisons; no prediction reduces to a fit or to a self-citation.

full rationale

The paper's central claim is that injecting an estimated blur kernel via the proposed Frequency Integration Module improves blind image deblurring and generalization. Walking the derivation chain: the kernel estimator is trained with the reblurring loss in Eq. (1), using ground-truth sharp images and input blurry images; the FIM is a newly proposed fusion module implemented with FFT/IFFT operations in Eqs. (2)-(4); and the final losses in Eqs. (6)-(7) supervise the deblurred image and reblur consistency. No reported test metric (GoPro, HIDE, RealBlur) is used as a fitting target, and no equation reduces to its own output by construction. The backbone [12] and kernel estimator [11] are external prior works with no author overlap with the present paper, and the paper explicitly states it retrained [12] and used its retrained result rather than the published number. The Table 3 ablation baseline (32.45 dB on GoPro) being 1.66 dB below the retrained FFTFormer baseline (about 34.11 dB) is an experimental-reporting inconsistency that weakens the attribution of the 0.74 dB gain to the kernel prior, but it is a comparison-protocol and reproducibility concern, not a circularity: the controlled rows within Table 3 are still trained under the same stated setting, and the non-blind oracle experiment in Table 4 provides an independent sanity check that accurate kernels help. Thus there is no circular step of the kind the analysis targets, and the honest finding is no significant circularity.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central claim depends on the standard blur formation model, the convolution theorem, and the assumption that per-pixel estimated kernels capture the degradation. Loss weights lambda1 and lambda2 are chosen by hand. No invented entities are introduced.

free parameters (2)
  • loss weight lambda1 (frequency loss) = 0.1
    Set in stage III (Sec 4.1); chosen by hand, not justified by analysis.
  • loss weight lambda2 (reblurring loss) = 0.1
    Set in stage III (Sec 4.1); chosen by hand, not justified by analysis.
assumptions (5)
  • domain assumption Blurry image y is generated as y = k ⊗ x + n, a convolution of a sharp image with a blur kernel plus noise.
    Stated in Sec 1 and used throughout the method.
  • standard math Convolution in the spatial domain equals pointwise multiplication in the frequency domain (deconvolution theorem).
    Invoked in Sec 3.2 to justify the FIM design.
  • domain assumption Non-uniform blur can be represented by a per-pixel kernel map estimated from the blurry image.
    Kernel estimation module in Sec 3.1 predicts Bn for each pixel, relying on the KoalaNet formulation [11].
  • domain assumption The reblurring loss ||B ⊗ x - y|| is a sufficient supervisory signal for kernel estimation.
    Used in Eq. (1) and Eq. (7); assumes the forward model is invertible enough for self-supervision.
  • domain assumption Fourier transforms of deep feature maps retain useful blur-related information that can be fused multiplicatively.
    Core design choice of FIM in Sec 3.2; not derived or formally justified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Frequency-domain Learning with Kernel Prior for Blind Image Deblurring." pith.science (2026). https://pith.science/paper/2FJBGNYG

@misc{pith2026250414664,
  author       = {Pith},
  title        = {Pith review of: Frequency-domain Learning with Kernel Prior for Blind Image Deblurring},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2FJBGNYG}},
  note         = {Machine review of arXiv:2504.14664}
}
read the original abstract

While achieving excellent results on various datasets, many deep learning methods for image deblurring suffer from limited generalization capabilities with out-of-domain data. This limitation is likely caused by their dependence on certain domain-specific datasets. To address this challenge, we argue that it is necessary to introduce the kernel prior into deep learning methods, as the kernel prior remains independent of the image context. For effective fusion of kernel prior information, we adopt a rational implementation method inspired by traditional deblurring algorithms that perform deconvolution in the frequency domain. We propose a module called Frequency Integration Module (FIM) for fusing the kernel prior and combine it with a frequency-based deblurring Transfomer network. Experimental results demonstrate that our method outperforms state-of-the-art methods on multiple blind image deblurring tasks, showcasing robust generalization abilities. Source code will be available soon.

Figures

Figures reproduced from arXiv: 2504.14664 by the authors.

Figure 1
Figure 1. Overview of the network architecture of our proposed method. (a) The structure of the Kernel Estimation Module. (b) The [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Visual comparison on GoPro [21] dataset. We compare our method with MIMO-UNet+ [7], Stripformer [32], Restormer￾Local [37], NAFNet [5], and FFTFormer [12]. Models are trained only on the GoPro dataset. Our network generates more realistic images with clearer details. image feature Xin ∈ R Cˆ×Hˆ ×Wˆ , and a blur kernel feature Bin ∈ R Dˆ×Hˆ ×Wˆ , where Hˆ ×Wˆ is the spatial resolution of input features, Cˆ and Dˆ are… view at source ↗
Figure 3
Figure 3. Visual comparison on RealBlur [26] dataset. We compare our method with MIMO-UNet+ [7], NAFNet [5], Restormer-Local [37], Stripformer [32], and FFTFormer [12]. Models are trained only on the GoPro dataset. Our network generates more realistic images with clearer details. a real-world dataset (RealBlur [26]). GoPro dataset in￾cludes 1,111 test images, HIDE dataset provides 2,025 im￾ages for testing, and RealBlur consi… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Visual comparison on HIDE [28] dataset. We compare our method with MIMO-UNet+ [7], NAFNet [5], Restormer-Local [37], Stripformer [32], and FFTFormer [12]. Models are trained only on the GoPro dataset. Our network generates more realistic images with clearer details. be…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 29 canonical work pages

  1. [1]

    Non-uniform blind deblurring by reblurring

    Yuval Bahat, Netalee Efrat, and Michal Irani. Non-uniform blind deblurring by reblurring. In Proceedings of the IEEE international conference on computer vision , pages 3286– 3294, 2017. 3

    work page 2017

  2. [2]

    Blind motion deblurring with pixel-wise kernel esti- mation via kernel prediction networks

    Guillermo Carbajal, Patricia Vitoria, Jos ´e Lezama, and Pablo Mus´e. Blind motion deblurring with pixel-wise kernel esti- mation via kernel prediction networks. IEEE Transactions on Computational Imaging, 2023. 1, 2

    work page 2023

  3. [3]

    A neural approach to blind motion deblur- ring

    Ayan Chakrabarti. A neural approach to blind motion deblur- ring. In Computer Vision–ECCV 2016: 14th European Con- ference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part III 14, pages 221–235. Springer, 2016. 1

    work page 2016

  4. [4]

    Reblur2deblur: Deblurring videos via self-supervised learning

    Huaijin Chen, Jinwei Gu, Orazio Gallo, Ming-Yu Liu, Ashok Veeraraghavan, and Jan Kautz. Reblur2deblur: Deblurring videos via self-supervised learning. In 2018 IEEE Interna- tional Conference on Computational Photography (ICCP) , pages 1–9. IEEE, 2018. 2

    work page 2018

  5. [5]

    Simple baselines for image restoration

    Liangyu Chen, Xiaojie Chu, Xiangyu Zhang, and Jian Sun. Simple baselines for image restoration. In European Confer- ence on Computer Vision, pages 17–33. Springer, 2022. 4, 5, 6, 7

    work page 2022

  6. [6]

    Hierarchical Integration Diffusion Model for Realistic Image Deblurring

    Zheng Chen, Yulun Zhang, Ding Liu, Bin Xia, Jinjin Gu, Linghe Kong, and Xin Yuan. Hierarchical integration dif- fusion model for realistic image deblurring. arXiv preprint arXiv:2305.12966, 2023. 5

    work page Pith review arXiv 2023

  7. [7]

    Rethinking coarse-to-fine approach in sin- gle image deblurring

    Sung-Jin Cho, Seo-Won Ji, Jun-Pyo Hong, Seung-Won Jung, and Sung-Jea Ko. Rethinking coarse-to-fine approach in sin- gle image deblurring. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 4641–4650,

  8. [8]

    Improving image restoration by revisiting global information aggregation

    Xiaojie Chu, Liangyu Chen, Chengpeng Chen, and Xin Lu. Improving image restoration by revisiting global information aggregation. In European Conference on Computer Vision, pages 53–71. Springer, 2022. 5, 7

    work page 2022

Show all 39 references
  1. [9]

    Dual- domain attention for image deblurring

    Yuning Cui, Yi Tao, Wenqi Ren, and Alois Knoll. Dual- domain attention for image deblurring. In Proceedings of the AAAI Conference on Artificial Intelligence , pages 479–487,

  2. [10]

    Deblurring using analysis-synthesis networks pair

    Adam Kaufman and Raanan Fattal. Deblurring using analysis-synthesis networks pair. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5811–5820, 2020. 1, 2

    2020

  3. [11]

    Koalanet: Blind super-resolution using kernel-oriented adaptive local adjustment

    Soo Ye Kim, Hyeonjun Sim, and Munchurl Kim. Koalanet: Blind super-resolution using kernel-oriented adaptive local adjustment. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 10611– 10620, 2021. 3

    2021

  4. [12]

    Efficient frequency domain-based trans- formers for high-quality image deblurring

    Lingshun Kong, Jiangxin Dong, Jianjun Ge, Mingqiang Li, and Jinshan Pan. Efficient frequency domain-based trans- formers for high-quality image deblurring. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 5886–5895, 2023. 2, 4, 5, 6, 7

    2023

  5. [13]

    Fast image deconvolution using hyper-laplacian priors

    Dilip Krishnan and Rob Fergus. Fast image deconvolution using hyper-laplacian priors. Advances in neural information processing systems, 22, 2009. 1, 2

    2009

  6. [14]

    Learning to push the limits of efficient fft-based image deconvolution

    Jakob Kruse, Carsten Rother, and Uwe Schmidt. Learning to push the limits of efficient fft-based image deconvolution. In Proceedings of the IEEE International Conference on Com- puter Vision, pages 4586–4594, 2017. 1, 2

    2017

  7. [15]

    Deblurgan-v2: Deblurring (orders-of-magnitude) faster and better

    Orest Kupyn, Tetiana Martyniuk, Junru Wu, and Zhangyang Wang. Deblurgan-v2: Deblurring (orders-of-magnitude) faster and better. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 8878–8887,

  8. [16]

    Understanding and evaluating blind deconvolution al- gorithms

    Anat Levin, Yair Weiss, Fredo Durand, and William T Free- man. Understanding and evaluating blind deconvolution al- gorithms. In 2009 IEEE conference on computer vision and pattern recognition, pages 1964–1971. IEEE, 2009. 3

    2009

  9. [17]

    Learning degradation rep- resentations for image deblurring

    Dasong Li, Yi Zhang, Ka Chun Cheung, Xiaogang Wang, Hongwei Qin, and Hongsheng Li. Learning degradation rep- resentations for image deblurring. In European Conference on Computer Vision, pages 736–753. Springer, 2022. 1, 2, 5

    2022

  10. [18]

    Effi- cient and explicit modelling of image hierarchies for image restoration

    Yawei Li, Yuchen Fan, Xiaoyu Xiang, Denis Demandolx, Rakesh Ranjan, Radu Timofte, and Luc Van Gool. Effi- cient and explicit modelling of image hierarchies for image restoration. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18278–...

    2023

  11. [19]

    Deep residual fourier transformation for single image deblurring

    Xintian Mao, Yiming Liu, Wei Shen, Qingli Li, and Yan Wang. Deep residual fourier transformation for single image deblurring. arXiv preprint arXiv:2111.11745, 2(3):5, 2021. 5, 7

    2021 arXiv

  12. [20]

    Intriguing findings of frequency selection for image deblurring

    Xintian Mao, Yiming Liu, Fengze Liu, Qingli Li, Wei Shen, and Yan Wang. Intriguing findings of frequency selection for image deblurring. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 1905–1913, 2023. 2

    1905

  13. [21]

    Deep multi-scale convolutional neural network for dynamic scene deblurring

    Seungjun Nah, Tae Hyun Kim, and Kyoung Mu Lee. Deep multi-scale convolutional neural network for dynamic scene deblurring. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3883–3891,

  14. [22]

    Region-adaptive dense network for efficient motion deblurring

    Kuldeep Purohit and AN Rajagopalan. Region-adaptive dense network for efficient motion deblurring. In Proceed- ings of the AAAI Conference on Artificial Intelligence, pages 11882–11889, 2020. 1

    2020

  15. [23]

    Neural blind deconvolution using deep pri- ors

    Dongwei Ren, Kai Zhang, Qilong Wang, Qinghua Hu, and Wangmeng Zuo. Neural blind deconvolution using deep pri- ors. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition , pages 3341–3350,

  16. [24]

    Video deblurring via semantic segmentation and pixel- wise non-linear kernel

    Wenqi Ren, Jinshan Pan, Xiaochun Cao, and Ming-Hsuan Yang. Video deblurring via semantic segmentation and pixel- wise non-linear kernel. In Proceedings of the IEEE Inter- national Conference on Computer Vision, pages 1077–1085,

  17. [25]

    Deep non-blind deconvolution via generalized low-rank approxi- mation

    Wenqi Ren, Jiawei Zhang, Lin Ma, Jinshan Pan, Xiaochun Cao, Wangmeng Zuo, Wei Liu, and Ming-Hsuan Yang. Deep non-blind deconvolution via generalized low-rank approxi- mation. Advances in neural information processing systems, 31, 2018. 2

    2018

  18. [26]

    Real-world blur dataset for learning and benchmarking de- blurring algorithms

    Jaesung Rim, Haeyun Lee, Jucheol Won, and Sunghyun Cho. Real-world blur dataset for learning and benchmarking de- blurring algorithms. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXV 16, pages 184–201. Springer, 20...

    2020

  19. [27]

    Learning to deblur

    Christian J Schuler, Michael Hirsch, Stefan Harmeling, and Bernhard Sch¨olkopf. Learning to deblur. IEEE transactions on pattern analysis and machine intelligence , 38(7):1439– 1451, 2015. 1, 2

    2015

  20. [28]

    Human-aware mo- tion deblurring

    Ziyi Shen, Wenguan Wang, Xiankai Lu, Jianbing Shen, Haibin Ling, Tingfa Xu, and Ling Shao. Human-aware mo- tion deblurring. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision , pages 5572–5581,

  21. [29]

    Spatially-attentive patch-hierarchical network for adaptive motion deblurring

    Maitreya Suin, Kuldeep Purohit, and AN Rajagopalan. Spatially-attentive patch-hierarchical network for adaptive motion deblurring. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 3606–3615, 2020. 5

    2020

  22. [30]

    Learn- ing a convolutional neural network for non-uniform motion blur removal

    Jian Sun, Wenfei Cao, Zongben Xu, and Jean Ponce. Learn- ing a convolutional neural network for non-uniform motion blur removal. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 769–777,

  23. [31]

    Scale-recurrent network for deep image deblurring

    Xin Tao, Hongyun Gao, Xiaoyong Shen, Jue Wang, and Ji- aya Jia. Scale-recurrent network for deep image deblurring. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 8174–8182, 2018. 5

    2018

  24. [32]

    Stripformer: Strip transformer for fast image deblurring

    Fu-Jen Tsai, Yan-Tsung Peng, Yen-Yu Lin, Chung-Chi Tsai, and Chia-Wen Lin. Stripformer: Strip transformer for fast image deblurring. In European Conference on Computer Vi- sion, pages 146–162. Springer, 2022. 4, 5, 6, 7

    2022

  25. [33]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Si- moncelli. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13(4):600–612, 2004. 6

    2004

  26. [34]

    Deep convolu- tional neural network for image deconvolution

    Li Xu, Jimmy S Ren, Ce Liu, and Jiaya Jia. Deep convolu- tional neural network for image deconvolution. Advances in neural information processing systems, 27, 2014. 2

    2014

  27. [35]

    Motion blur kernel estimation via deep learning.IEEE Transactions on Image Processing, 27(1):194–205, 2017

    Xiangyu Xu, Jinshan Pan, Yu-Jin Zhang, and Ming-Hsuan Yang. Motion blur kernel estimation via deep learning.IEEE Transactions on Image Processing, 27(1):194–205, 2017. 2

    2017

  28. [36]

    Multi-stage progressive image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan, Ming-Hsuan Yang, and Ling Shao. Multi-stage progressive image restoration. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14821–14831, 2021. 5

    2021

  29. [37]

    Restormer: Efficient transformer for high-resolution image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Mu- nawar Hayat, Fahad Shahbaz Khan, and Ming-Hsuan Yang. Restormer: Efficient transformer for high-resolution image restoration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5728–5739,

  30. [38]

    Deep stacked hierarchical multi-patch network for im- age deblurring

    Hongguang Zhang, Yuchao Dai, Hongdong Li, and Piotr Ko- niusz. Deep stacked hierarchical multi-patch network for im- age deblurring. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 5978– 5986, 2019. 5

    2019

  31. [39]

    From learning models of nat- ural image patches to whole image restoration

    Daniel Zoran and Yair Weiss. From learning models of nat- ural image patches to whole image restoration. In 2011 in- ternational conference on computer vision , pages 479–486. IEEE, 2011. 2

    2011

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.