Pith. sign in

REVIEW 3 major objections 7 minor 53 references

Complementary Advantages: Exploiting Cross-Field Frequency Correlation for NIR-Assisted Image Denoising

T0 review · 3 major / 7 minor · reviewed 2026-08-11 · 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 prior for NIR-assisted denoising: keep RGB low-frequency color, take NIR high-frequency detail.

desk verdict Solid NIR-assisted denoising paper with a real empirical prior, but the fusion module's equations are underspecified and need a fix before I'd trust the architecture as written. read the letter →

arxiv 2412.16645 v2 pith:HADZ6K53 submitted 2024-12-21 cs.CV

Yuchen Wang , Hongyuan Wang , Lizhi Wang , Xin Wang , Lin Zhu , Wanxuan Lu , Hua Huang This is my paper · ORCID
classification cs.CV
keywords imagedenoisingnear-infraredimagingfrequencydomainanalysiscross-fieldfusiondynamicconvolutionattentionmechanismguidedrestoration
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

The paper establishes a frequency-correlation prior for NIR-assisted RGB denoising: in the same scene, noisy RGB images become progressively less similar to the clean image from low to high frequency, while NIR images become progressively more similar. The authors argue that this complementary relationship makes frequency-domain selection the right way to fuse the two fields. They build FCENet, which uses a Frequency Dynamic Selection Mechanism to filter useful information from each field and a Frequency Exhaustive Fusion Mechanism to combine common and differential features. On the DVD and IVRG datasets, FCENet reports higher PSNR and SSIM than prior single-image and NIR-assisted denoisers, including a light version with only 1.16 million parameters.

What carries the argument

The Frequency Dynamic Selection Mechanism (FDSM) generates input-dependent frequency-domain filters by learning linear combinations of k learnable filter kernels from aggregated NIR and RGB features, then applies these filters via element-wise multiplication in the 2D-DFT domain. The Frequency Exhaustive Fusion Mechanism (FEFM) has two branches: Common Feature Reinforcement (CFR) builds a frequency-domain correlation map that combines point-wise products with long-range channel correlations, and Differential Feature Reinforcement (DFR) uses a differential cross-attention to add missing high-frequency NIR features back into the fused representation. The whole network is a two-stage U-Net with a Supervised Attention Module connecting the pre-denoised RGB stage to the NIR-fusion stage.

What would settle it

Compute the SSIM-versus-cutoff-frequency curves on the full DVD and IVRG test sets, or on a new NIR-RGB dataset, with per-frequency error bars. If the noisy-RGB curve does not consistently fall while the NIR curve rises, or if the ordering reverses at any band, the frequency-correlation prior is contradicted and the FDSM/FEFM design loses its motivating premise.

Watch

Extended reading notes

Core claim

The central discovery is the cross-field frequency correlation prior, measured by applying high-pass filters at increasing cutoff frequencies to 30 randomly selected NIR-RGB image pairs and computing SSIM between the filtered targets and the clean RGB image. The paper finds that for noisy RGB, SSIM with clean RGB falls as frequency rises, while for NIR the SSIM with clean RGB rises. The authors then claim that this prior justifies a frequency-domain fusion framework: FDSM learns input-dependent frequency filters that retain low-frequency color from RGB and mid-to-high-frequency texture from NIR, and FEFM reinforces common features through a frequency-domain correlation map while supplementing differential high-frequency features from NIR via differential cross-attention. With this design, FCENet reaches 32.43 dB PSNR on DVD at noise level σ=2 and 33.26 dB on IVRG at σ=25, outperforming all compared methods.

Load-bearing premise

The entire fusion scheme rests on the measured trend that noisy RGB similarity to clean decreases with frequency while NIR similarity increases, yet the trend is computed on only 30 image pairs with no reported variance or statistical test.

Editorial extensions

If this is right

  • If the frequency-correlation prior holds broadly, NIR-assisted denoising is better formulated as frequency-domain selection than as spatial fusion, which should reduce artifacts caused by color and structure inconsistency between fields.
  • Restoring high-frequency texture from NIR while preserving low-frequency color from RGB yields higher PSNR and SSIM than single-image denoisers or additive fusion, as shown on DVD and IVRG.
  • Dynamic frequency-domain filtering is parameter-efficient: FCENet-L with 1.16 million parameters outperforms heavier baselines.
  • The two-stage design, where RGB is pre-denosed before fusion, reduces the modality gap and makes subsequent frequency fusion more effective.

Reading between the lines

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

  • The paper does not claim, but it follows that the same frequency-selection structure could transfer to other guided restoration tasks, such as RGB-depth super-resolution or flash/no-flash photography, where one modality supplies high-frequency structure and the other supplies low-frequency color.
  • Because the prior is measured on 30 pairs with no error bars, a direct stress test would be to recompute the SSIM-vs-frequency curves separately per noise level and per scene; if the NIR curve does not consistently rise, the design's motivation weakens.
  • The paper admits that significantly misaligned cross-field images break existing methods; a testable extension is to insert an alignment module before FCENet and measure whether the frequency-selection gains are preserved under realistic misalignment.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 7 minor

Summary. The paper proposes FCENet, a two-stage U-Net architecture for NIR-assisted RGB image denoising. The first stage pre-denoises the RGB image; the second stage fuses NIR and RGB features using two newly proposed modules: Frequency Dynamic Selection Mechanism (FDSM) and Frequency Exhaustive Fusion Mechanism (FEFM). The design is motivated by a claimed frequency correlation prior: across the same scene, noisy RGB becomes less similar to clean RGB at higher frequencies, while NIR becomes more similar to clean RGB at higher frequencies. The authors evaluate on the DVD and IVRG datasets under several noise levels, report PSNR/SSIM improvements over existing NIR-assisted and single-image denoising methods, and provide ablations showing that FDSM and FEFM each contribute to the final performance.

Significance. If the results hold, the paper would make a useful contribution to NIR-assisted denoising by reframing cross-field fusion as frequency-domain selection and fusion. The experimental setup is a strength: all compared methods are retrained on the same training sets, the ablations in Tables 3-5 show that each proposed component contributes, and the method reports favorable FLOPs/parameter trade-offs. The frequency correlation prior is presented as a data-derived observation rather than a fitted parameter, so there is no obvious circularity in the evaluation. However, the central architectural equations for FEFM are not mathematically well-defined as written, and the prior is supported by only 30 image pairs without statistical quantification; both issues directly affect the reproducibility and credibility of the central claims.

major comments (3)
  1. [Sec. 3.4, Eqs. (4)-(5)] The core FEFM equations are underspecified: FQ and FK are 2D-DFTs of real feature maps and are therefore complex-valued, so Eq. (4) applies Softmax to a complex matrix and computes a Hadamard product of complex matrices without any stated real-valued reduction. Eq. (5) then forms V - λ V ⊙ F^{-1}(FCFR), which is complex-valued, yet the subsequent convolutional and residual operations require real feature maps. Since CFR and DFR are the two headline components of FEFM, and Table 3 attributes a 0.83 dB gain to FEFM, the published description cannot be implemented or reproduced as written. Please specify the complex-to-real mapping (e.g., taking the real part, magnitude, or a learned projection), define the dimensions of the matrix products in Eq. (4), and clarify how the final fused result combines Q with the output of CFR.
  2. [Sec. 3.1, Fig. 3] The frequency correlation prior is the load-bearing motivation for both FDSM and FEFM, but it is established from only 30 randomly selected image pairs, with no error bars, confidence intervals, or statistical test. Fig. 3 shows individual curves rather than an aggregate summary, so the claimed monotonic trends are not demonstrated to hold across the full training and test distributions, different noise levels, or different sensor/scene types. Please provide a quantitative summary over the full DVD and IVRG sets (e.g., mean ± std of the SSIM-frequency curves), state which dataset the 30 pairs were drawn from, and ideally test the monotonicity claim statistically.
  3. [Tables 1-2 and Sec. 4.2] The claim that FCENet outperforms all compared methods rests on single-run PSNR/SSIM differences, some of which are small (e.g., Table 1, σ=6: FCENet-L 28.36 vs. NIR-Restormer 28.31; Table 2, σ=75: FCENet-L 25.33 vs. NIR-Restormer 25.30). Without multiple seeds, error bars, or significance tests, it is difficult to judge whether these differences are beyond run-to-run variation. Please report means and standard deviations over at least three training runs, or otherwise justify that the reported single-run differences are stable.
minor comments (7)
  1. [Sec. 4.2] The subsection heading "Evaluations on the DVN dataset" appears to be a typo for "DVD dataset."
  2. [Eq. (1)] The high-pass filter H(·) is not precisely defined; please specify whether it is a radial frequency mask, how the cutoff frequency is normalized, and how the filtered frequency-domain representation is converted back to the spatial domain.
  3. [Sec. 3.3, Eq. (2)] In Eq. (2), G is described as containing k learnable frequency-domain filtering kernels, but it is not stated whether DFI and G are real-valued or complex-valued. Since F(I) in Eq. (3) is complex, please clarify the domain of the dynamic filter and the intended multiplication.
  4. [Table 3] The header of Table 3 is confusing because the first column is also labeled "Baseline" and row (a) has a checkmark only under that column; please rename the columns to explicitly indicate the configuration (e.g., "FDSM" and "FEFM" with a separate "w/o modules" column).
  5. [Sec. 4.1 and Table 1] The training noise specification says Gaussian-Poisson mixed noise with a range of 1 to 16, while Table 1 evaluates at σ=2, 4, and 6; please explain how these test noise levels relate to the training noise range and whether the model sees the exact test noise levels during training.
  6. [Sec. 5, Conclusion] The final paragraph notes that existing methods may fail on significantly misaligned cross-field images and lists this as future work; please state explicitly whether FCENet inherits this limitation and confirm that the DVD and IVRG evaluation sets do not contain significant misalignment.
  7. [Fig. 5] The notation in Fig. 5 (QF, KF, V) does not fully match the text in Sec. 3.4 (Q, K, V, FQ, FK); please align the notations for readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the frequency correlation prior is an empirical observation and the network is validated against external benchmarks.

full rationale

The paper's central claim is the frequency correlation prior, obtained by measuring SSIM between high-pass-filtered NIR/noisy-RGB and clean-RGB images on 30 randomly selected pairs (Eq. 1, Sec. 3.1). This is an empirical observation, not a fitted parameter of the network, and it is not derived from the network's outputs; the prior only motivates the FDSM/FEFM architecture. Performance is established by training on the DVD and IVRG training splits and testing on held-out splits (Tables 1-2), with all compared methods re-trained on the same training set, so the reported PSNR/SSIM gains are external and falsifiable rather than forced by construction. The ablation studies (Tables 3-5) compare architectural variants against a baseline under identical protocol, and the gains are empirical. Several references (e.g., [13], [14], [22], [23], [39]) involve the present authors, but none is load-bearing for the frequency prior or the architecture's motivation; no uniqueness theorem or ansatz is imported from self-cited prior work. The conclusion honestly states that existing methods—including the proposed approach—may fail on significantly misaligned cross-field images, which is a scope limitation, not circularity. The skeptic's objection that Eqs. (4)-(5) apply Softmax and subtraction to complex-valued frequency-domain tensors without specifying a real-valued reduction is a mathematical specification and reproducibility concern, not a circular derivation; under the review rules, such concerns belong to correctness risk, not circularity. No circular step can be exhibited, so the appropriate finding is no significant circularity.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The method rests on standard deep learning components and a new empirical prior. The only hand-set scalar is the loss weight; the network's learned weights are not free parameters in the derivation sense. The main input-dependent assumption is that the frequency correlation prior holds broadly across the datasets.

free parameters (1)
  • Loss weight alpha (Eq. 6) = 0.1
    Hand-chosen scalar controlling the frequency-domain loss contribution; taken from prior work, not central to the method's novelty.
assumptions (3)
  • domain assumption The DVD and IVRG datasets provide representative aligned NIR-RGB pairs typical of NIR-assisted denoising.
    The network is trained and evaluated on these datasets; if they are not representative of real-world misalignment, the method's advantage may not transfer.
  • domain assumption The two-stage pre-denoising is sufficient to reduce the modality gap so that encoder features of NIR and RGB can be fused in frequency domain.
    The entire fusion design (FDSM/FEFM) operates on features after pre-denoising; if the pre-denoised RGB retains too much noise or loses texture, the frequency fusion is built on poor inputs.
  • standard math Standard FFT and convolution theorem properties are used without proof.
    Eqs. 3 and 4 rely on 2D-DFT/IDFT and the convolution theorem.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Complementary Advantages: Exploiting Cross-Field Frequency Correlation for NIR-Assisted Image Denoising." pith.science (2026). https://pith.science/paper/HADZ6K53

@misc{pith2026241216645,
  author       = {Pith},
  title        = {Pith review of: Complementary Advantages: Exploiting Cross-Field Frequency Correlation for NIR-Assisted Image Denoising},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HADZ6K53}},
  note         = {Machine review of arXiv:2412.16645}
}
read the original abstract

Existing single-image denoising algorithms often struggle to restore details when dealing with complex noisy images. The introduction of near-infrared (NIR) images offers new possibilities for RGB image denoising. However, due to the inconsistency between NIR and RGB images, the existing works still struggle to balance the contributions of two fields in the process of image fusion. In response to this, in this paper, we develop a cross-field Frequency Correlation Exploiting Network (FCENet) for NIR-assisted image denoising. We first propose the frequency correlation prior based on an in-depth statistical frequency analysis of NIR-RGB image pairs. The prior reveals the complementary correlation of NIR and RGB images in the frequency domain. Leveraging frequency correlation prior, we then establish a frequency learning framework composed of Frequency Dynamic Selection Mechanism (FDSM) and Frequency Exhaustive Fusion Mechanism (FEFM). FDSM dynamically selects complementary information from NIR and RGB images in the frequency domain, and FEFM strengthens the control of common and differential features during the fusion process of NIR and RGB features. Extensive experiments on simulated and real data validate that the proposed method outperforms other state-of-the-art methods. The code will be released at https://github.com/yuchenwang815/FCENet.

Figures

Figures reproduced from arXiv: 2412.16645 by the authors.

Figure 1
Figure 1. Comparisons of PSNR, FLOPs, and Parameters on the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Visual comparisons on the challenging noisy RGB-NIR [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The analysis of NIR and RGB frequency correlation. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: The overall architecture of the proposed cross-field Fre [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: The architecture of the proposed Frequency Dynamic Selection Mechanism (FDSM) and Frequency Exhaustive Fusion Mecha [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: The qualitative comparison among our FCENet and the state-of-the-art methods on the noisy RGB-NIR pairs from DVD[ [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Visual comparisons on real-world RGB/NIR image pairs. Our method demonstrates better visual results. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Frequency domain dynamic filters on NIR and RGB side [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: Feature maps obtained from two fields. The first column [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 46 canonical work pages

  1. [1]

    Ntire 2020 challenge on real image denoising: Dataset, methods and results

    Abdelrahman Abdelhamed, Mahmoud Afifi, Radu Timofte, and Michael S Brown. Ntire 2020 challenge on real image denoising: Dataset, methods and results. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, pages 496–497, 2020. 1

    work page 2020

  2. [2]

    Real image denoising with feature attention

    Saeed Anwar and Nick Barnes. Real image denoising with feature attention. In Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 3155–3164,

    work page

  3. [3]

    Noise2self: Blind denoising by self-supervision

    Joshua Batson and Loic Royer. Noise2self: Blind denoising by self-supervision. In International Conference on Machine Learning, pages 524–533, 2019. 1

    work page 2019

  4. [4]

    Multi-spectral sift for scene category recognition

    Matthew Brown and Sabine S ¨usstrunk. Multi-spectral sift for scene category recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 177– 184, 2011. 6

    work page 2011

  5. [5]

    Two deterministic half-quadratic regular- ization algorithms for computed imaging

    Pierre Charbonnier, Laure Blanc-Feraud, Gilles Aubert, and Michel Barlaud. Two deterministic half-quadratic regular- ization algorithms for computed imaging. In Proceedings of the International Conference on Image Processing, pages 168–172, 1994. 5

    work page 1994

  6. [6]

    Dynamic convolution: Attention over convolution kernels

    Yinpeng Chen, Xiyang Dai, Mengchen Liu, Dongdong Chen, Lu Yuan, and Zicheng Liu. Dynamic convolution: Attention over convolution kernels. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11030–11039, 2020. 4

    work page 2020

  7. [7]

    Nbnet: Noise basis learning for image denoising with subspace projection

    Shen Cheng, Yuzhi Wang, Haibin Huang, Donghao Liu, Haoqiang Fan, and Shuaicheng Liu. Nbnet: Noise basis learning for image denoising with subspace projection. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 4896–4906, 2021. 2

    work page 2021

  8. [8]

    A mutually boosting dual sensor compu- tational camera for high quality dark videography

    Yuxiao Cheng, Runzhao Yang, Zhihong Zhang, Jinli Suo, and Qionghai Dai. A mutually boosting dual sensor compu- tational camera for high quality dark videography. Informa- tion Fusion, 93:429–440, 2023. 2

    work page 2023

Show all 53 references
  1. [9]

    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. InProceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 4641–4650,

  2. [10]

    Selective frequency network for image restoration

    Yuning Cui, Yi Tao, Zhenshan Bing, Wenqi Ren, Xinwei Gao, Xiaochun Cao, Kai Huang, and Alois Knoll. Selective frequency network for image restoration. In International Conference on Learning Representations, 2023. 3

    2023

  3. [11]

    Deep convolutional neural network for multi-modal image restoration and fusion

    Xin Deng and Pier Luigi Dragotti. Deep convolutional neural network for multi-modal image restoration and fusion. IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(10):3333–3348, 2020. 2, 3, 6, 7

    2020

  4. [12]

    Image denoising via sparse and redundant representations over learned dictionar- ies

    Michael Elad and Michal Aharon. Image denoising via sparse and redundant representations over learned dictionar- ies. IEEE Transactions on Image processing, 15(12):3736– 3745, 2006. 2

    2006

  5. [13]

    Learnability enhancement for low-light raw denoising: Where paired real data meets noise modeling

    Hansen Feng, Lizhi Wang, Yuzhi Wang, and Hua Huang. Learnability enhancement for low-light raw denoising: Where paired real data meets noise modeling. In Proceed- ings of the ACM International Conference on Multimedia , pages 1436–1444, 2022. 1

    2022

  6. [14]

    Learnability enhancement for low-light raw im- age denoising: A data perspective

    Hansen Feng, Lizhi Wang, Yuzhi Wang, Haoqiang Fan, and Hua Huang. Learnability enhancement for low-light raw im- age denoising: A data perspective. IEEE Transactions on Pattern Analysis and Machine Intelligence , 46(1):370–387,

  7. [15]

    Mutual-guided dynamic network for image fusion

    Yuanshen Guan, Ruikang Xu, Mingde Yao, Lizhi Wang, and Zhiwei Xiong. Mutual-guided dynamic network for image fusion. In Proceedings of the ACM International Conference on Multimedia, pages 1779–1788, 2023. 5

    2023

  8. [16]

    Hd- net: High-resolution dual-domain learning for spectral com- pressive imaging

    Xiaowan Hu, Yuanhao Cai, Jing Lin, Haoqian Wang, Xin Yuan, Yulun Zhang, Radu Timofte, and Luc Van Gool. Hd- net: High-resolution dual-domain learning for spectral com- pressive imaging. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pag...

    2022

  9. [17]

    Focal frequency loss for image reconstruction and synthesis

    Liming Jiang, Bo Dai, Wayne Wu, and Chen Change Loy. Focal frequency loss for image reconstruction and synthesis. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 13919–13929, 2021. 3

    2021

  10. [18]

    Darkvisionnet: Low-light imaging via rgb-nir fusion with deep inconsistency prior

    Shuangping Jin, Bingbing Yu, Minhao Jing, Yi Zhou, Jia- jun Liang, and Renhe Ji. Darkvisionnet: Low-light imaging via rgb-nir fusion with deep inconsistency prior. InProceed- ings of the AAAI Conference on Artificial Intelligence, pages 1104–1112, 2022. 1, 2, 3, 6, 7

    2022

  11. [19]

    Deformable kernel networks for joint image filtering

    Beomjun Kim, Jean Ponce, and Bumsub Ham. Deformable kernel networks for joint image filtering. International Jour- nal of Computer Vision, 129(2):579–600, 2021. 2

    2021

  12. [20]

    Adam: A method for stochastic optimization

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 ,

    arXiv

  13. [21]

    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. 3, 5

    2023

  14. [22]

    Stimulating the diffusion model for image denoising via adaptive embedding and ensembling

    Tong Li, Hansen Feng, Lizhi Wang, Zhiwei Xiong, and Hua Huang. Stimulating the diffusion model for image denoising via adaptive embedding and ensembling. In IEEE Transac- tions on Pattern Analysis and Machine Intelligence , pages 8240–8257, 2024. 2

    2024

  15. [23]

    Positive2negative: Breaking the information-lossy barrier in self-supervised single image de- noising

    Tong Li, Lizhi Wang, Zhiyuan Xu, Lin Zhu, Wanxuan Lu, and Hua Huang. Positive2negative: Breaking the information-lossy barrier in self-supervised single image de- noising. Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition, 2025. 2

    2025

  16. [24]

    Swinir: Image restoration us- ing swin transformer

    Jingyun Liang, Jiezhang Cao, Guolei Sun, Kai Zhang, Luc Van Gool, and Radu Timofte. Swinir: Image restoration us- ing swin transformer. InProceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 1833–1844,

  17. [25]

    Cat: Cross attention in vision transformer

    Hezheng Lin, Xing Cheng, Xiangyu Wu, and Dong Shen. Cat: Cross attention in vision transformer. In Proceedings of the IEEE International Conference on Multimedia and Expo, pages 1–6, 2022. 5

    2022

  18. [26]

    An integrated enhancement solution for 24-hour colorful imag- ing

    Feifan Lv, Yinqiang Zheng, Yicheng Li, and Feng Lu. An integrated enhancement solution for 24-hour colorful imag- ing. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 11725–11732, 2020. 1, 2

    2020

  19. [27]

    Gen- erative adaptive convolutions for real-world noisy image de- noising

    Ruijun Ma, Shuyi Li, Bob Zhang, and Zhengming Li. Gen- erative adaptive convolutions for real-world noisy image de- noising. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 1935–1943, 2022. 4

    1935

  20. [28]

    Zero-shot noise2noise: Efficient image denoising without any data

    Youssef Mansour and Reinhard Heckel. Zero-shot noise2noise: Efficient image denoising without any data. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 14018–14027, 2023. 1

    2023

  21. [29]

    Image denoising using scale mixtures of gaussians in the wavelet domain

    Javier Portilla, Vasily Strela, Martin J Wainwright, and Eero P Simoncelli. Image denoising using scale mixtures of gaussians in the wavelet domain. IEEE Transactions on Image Processing, 12(11):1338–1351, 2003. 2

    2003

  22. [30]

    U-net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Inter- vention, pages 234–241, 2015. 4

    2015

  23. [31]

    Frequency-domain deep guided image de- noising

    Zehua Sheng, Xiongwei Liu, Si-Yuan Cao, Hui-Liang Shen, and Huaqi Zhang. Frequency-domain deep guided image de- noising. IEEE Transactions on Multimedia, 25:6767–6781,

  24. [32]

    Structure aggre- gation for cross-spectral stereo image guided denoising

    Zehua Sheng, Zhu Yu, Xiongwei Liu, Si-Yuan Cao, Yuqi Liu, Hui-Liang Shen, and Huaqi Zhang. Structure aggre- gation for cross-spectral stereo image guided denoising. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 13997–14006, 2023....

    2023

  25. [33]

    Fourier features let networks learn high frequency functions in low dimen- sional domains

    Matthew Tancik, Pratul Srinivasan, Ben Mildenhall, Sara Fridovich-Keil, Nithin Raghavan, Utkarsh Singhal, Ravi Ra- mamoorthi, Jonathan Barron, and Ren Ng. Fourier features let networks learn high frequency functions in low dimen- sional domains. Advances in Neural Information ...

    2020

  26. [34]

    Fft-based dynamic token mixer for vision

    Yuki Tatsunami and Masato Taki. Fft-based dynamic token mixer for vision. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 15328–15336, 2024. 4

    2024

  27. [35]

    Deep learning on image denoising: An overview

    Chunwei Tian, Lunke Fei, Wenxian Zheng, Yong Xu, Wang- meng Zuo, and Chia-Wen Lin. Deep learning on image denoising: An overview. Neural Networks, 131:251–275,

  28. [36]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in Neural Information Processing Systems, 30, 2017. 1

    2017

  29. [37]

    Renjie Wan, Boxin Shi, Wenhan Yang, Bihan Wen, Ling- Yu Duan, and Alex C. Kot. Purifying low-light images via near-infrared enlightened image. IEEE Transactions on Mul- timedia, 25:8006–8019, 2023. 1, 3

    2023

  30. [38]

    Stereoscopic dark flash for low-light photography

    Jian Wang, Tianfan Xue, Jonathan T Barron, and Jiawen Chen. Stereoscopic dark flash for low-light photography. In IEEE International Conference on Computational Photogra- phy, pages 1–10, 2019. 2

    2019

  31. [39]

    In2set: Intra-inter similarity exploit- ing transformer for dual-camera compressive hyperspectral imaging

    Xin Wang, Lizhi Wang, Xiangtian Ma, Maoqing Zhang, Lin Zhu, and Hua Huang. In2set: Intra-inter similarity exploit- ing transformer for dual-camera compressive hyperspectral imaging. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 24...

    2024

  32. [40]

    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 Process- ing, 13(4):600–612, 2004. 3

    2004

  33. [41]

    Uformer: A general u-shaped transformer for image restoration

    Zhendong Wang, Xiaodong Cun, Jianmin Bao, Wengang Zhou, Jianzhuang Liu, and Houqiang Li. Uformer: A general u-shaped transformer for image restoration. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 17683–17693, 2022. 6, 7

    2022

  34. [42]

    Unsupervised real-world image super resolution via domain-distance aware training

    Yunxuan Wei, Shuhang Gu, Yawei Li, Radu Timofte, Long- cun Jin, and Hengjie Song. Unsupervised real-world image super resolution via domain-distance aware training. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13385–13394, 2021. 3

    2021

  35. [43]

    Nir-assisted image denoising: A selective fusion ap- proach and a real-world benchmark datase

    Rongjian Xu, Zhilu Zhang, Renlong Wu, and Wangmeng Zuo. Nir-assisted image denoising: A selective fusion ap- proach and a real-world benchmark datase. arXiv preprint arXiv:2404.08514, 2024. 2, 3, 6, 7

    2024 arXiv

  36. [44]

    A model-driven net- work for guided image denoising

    Shuang Xu, Jiangshe Zhang, Jialin Wang, Kai Sun, Chunxia Zhang, Junmin Liu, and Junying Hu. A model-driven net- work for guided image denoising. Information Fusion, 85: 60–71, 2022. 2, 3, 6, 7

    2022

  37. [45]

    Frequency principle: Fourier analy- sis sheds light on deep neural networks

    Zhi-Qin John Xu, Yaoyu Zhang, Tao Luo, Yanyang Xiao, and Zheng Ma. Frequency principle: Fourier analy- sis sheds light on deep neural networks. arXiv preprint arXiv:1901.06523, 2019. 3

    1901 arXiv

  38. [46]

    Cross-field joint image restoration via scale map

    Qiong Yan, Xiaoyong Shen, Li Xu, Shaojie Zhuo, Xiaopeng Zhang, Liang Shen, and Jiaya Jia. Cross-field joint image restoration via scale map. In Proceedings of the IEEE Inter- national Conference on Computer Vision, pages 1537–1544,

  39. [47]

    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. 1, 2, 4, 6, 7

    2021

  40. [48]

    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– 573...

    2022

  41. [49]

    Xformer: Hybrid x- shaped transformer for image denoising

    Jiale Zhang, Yulun Zhang, Jinjin Gu, Jiahua Dong, Linghe Kong, and Xiaokang Yang. Xformer: Hybrid x- shaped transformer for image denoising. arXiv preprint arXiv:2303.06440, 2023. 2, 6, 7

    2023 arXiv

  42. [50]

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

    Kai Zhang, Wangmeng Zuo, Yunjin Chen, Deyu Meng, and Lei Zhang. Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising. IEEE Transactions on Image Processing, 26(7):3142–3155, 2017. 2

    2017

  43. [51]

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

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

  44. [52]

    Decoupled dynamic filter networks

    Jingkai Zhou, Varun Jampani, Zhixiong Pi, Qiong Liu, and Ming-Hsuan Yang. Decoupled dynamic filter networks. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 6647–6656, 2021. 4

    2021

  45. [53]

    Sdwnet: A straight dilated net- work with wavelet transformation for image deblurring

    Wenbin Zou, Mingchao Jiang, Yunchen Zhang, Liang Chen, Zhiyong Lu, and Yi Wu. Sdwnet: A straight dilated net- work with wavelet transformation for image deblurring. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 1895–1904, 2021. 3

    1904

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.