Pith. sign in

REVIEW 4 major objections 6 minor 49 references

Hybrid-Domain Synergistic Transformer for Hyperspectral Image Denoising

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A hyperspectral denoising network that routes features through an FFT-based gating unit and multiscale pyramids reports 0.94 dB PSNR over its spatial-spectral transformer baseline on real noisy images.

desk verdict Incremental but honest work-in-progress: the architecture and ablation order are plausible, but the headline gains rest on single runs and the paper itself says it is unfinished. read the letter →

arxiv 2507.20099 v1 pith:24KYAUCH submitted 2025-07-27 cs.CV

Haoyue Li (1) , Di Wu (1) ((1) School of Optoelectronic Science , Engineering , Soochow University) This is my paper · ORCID
classification cs.CV
keywords hyperspectralimagedenoisingfrequencydomaintransformeratrousspatialpyramidpoolingcross-domainattentiongatingmechanismmultiscaleconvolutionspatial-spectralmodeling
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 claims that hyperspectral image denoising improves when a spatial-spectral transformer is augmented with a frequency-domain branch that decouples noise from signal and a multiscale branch that captures local detail and global statistics. The proposed network, HDST, builds on the SERT transformer and adds an FFT-based preprocessing stage, a gated cross-domain attention that lets spatial features query frequency-domain noise patterns, and a multiscale atrous pyramid. On the Realistic benchmark it reports 30.62 dB PSNR against 29.68 dB for its baseline, a 0.94 dB gain, and on the ICVL synthetic benchmark a maximum 0.52 dB gain, largest under mixture noise. The paper argues that the improvement comes from a closed loop of spatial feature extraction, frequency-domain noise decoupling, and cross-domain calibration, not from any single module. A sympathetic reader would care because real HSI noise is spatially non-uniform and coupled across bands, a setting where RGB-oriented denoisers and pure spatial models tend to blur weak spectral features.

What carries the argument

The central mechanism is the frequency post-processing (FPP) unit inserted into the deepest residual transformer layers. Its forward pass is $F = \mathrm{FFT}(S)$, $F_c = \mathrm{Concat}(\mathrm{Re}\,F, \mathrm{Im}\,F)$, $F_{\mathrm{proc}} = \mathrm{ASPP\text{-}FFT}(F_c)$ with dilation rates $[2,4,8]$, $S' = \mathrm{Conv}_{3\times3}(\mathrm{IFFT}(F_{\mathrm{proc}}))$, $\mathrm{Gate} = \sigma(\mathrm{Conv}_{1\times1}(F_{\mathrm{proc}}))$, and fusion $F' = S \odot \mathrm{Gate} + \alpha\, S' \odot (1 - \mathrm{Gate})$. The gate is intended to keep clean areas ($\mathrm{Gate} \to 1$) and replace noisy regions ($\mathrm{Gate} \to 0$), so the frequency branch injects noise-prior information only where needed. A frequency-spatial collaborative attention module then uses spatial windows as queries against frequency-derived keys and values, and a learnable coefficient $\beta$ scales the residual. The HDMS multiscale module, applied after each transformer block with dilation rates $[2,4,8]$, is the second pillar: it supplies edge-to-global spatial features that give the frequency pyramid content to act on.

What would settle it

Train SERT and HDST on the same Realistic split with at least five random seeds and compute the 95% confidence interval of the per-image PSNR difference; if the interval contains zero, the claimed 0.94 dB improvement is not statistically established. As a second check, run HDST on pure i.i.d. Gaussian noise at several levels: the paper's own mechanism predicts little or no gain there, so a large gain would contradict the proposed frequency-gating explanation.

Watch

Extended reading notes

Core claim

HDST's central claim is that the coupled spatial-spectral noise of hyperspectral images can be separated by routing features through a frequency-domain post-processing unit and recombining them with spatial features under learned control. The denoiser keeps SERT's rectangular-window attention and spectral enhancement, then inserts, in the deepest transformer layers, an FFT-scale gated fusion module that transforms features with FFT, processes the concatenated real and imaginary parts with atrous convolutions at dilation rates 2, 4, and 8, and fuses the reconstructed spatial signal with the original using a learned gate and a scalar intensity parameter. A frequency-spatial collaborative attention module then treats spatial features as queries and frequency-domain features as keys and values, so spatial positions can retrieve the noise pattern that corrupts them, and a learnable coefficient controls the residual. The paper reports that the complete model reaches 30.62 dB PSNR on the Realistic dataset versus 29.68 dB for the SERT baseline, outperforms the baseline on most synthetic noise patterns, and that the ablation attributes the largest single gain to the multiscale pyramid while the full gain comes from the combination.

Load-bearing premise

The central evidence rests on single-run comparisons: each configuration is trained once, and the reported gains, especially the 0.05 to 0.15 dB differences in ablation steps and some synthetic noise types, are treated as meaningful without error bars or multi-seed statistics.

Editorial extensions

If this is right

  • On the Realistic benchmark, HDST improves PSNR from 29.68 to 30.62 dB over SERT, with SSIM rising from 0.9533 to 0.9555 and SAM falling from 2.536 to 2.417.
  • On the ICVL synthetic benchmark, HDST beats SERT by up to 0.52 dB under Gaussian+Mixture noise (40.52 vs 40.00 dB) and improves most other noise patterns, with the smallest gains under Non-i.i.d Gaussian.
  • The ablation shows the full gain is synergistic: multiscale-only gives +0.59 dB, frequency-plus-multiscale without dynamic fusion gives +0.78 dB, and the complete model gives +0.94 dB, while frequency processing alone gives only +0.05 dB.
  • The architecture adds 61.2% parameters but only 14.7% FLOPs, so the frequency and multiscale branches are claimed to be computationally inexpensive relative to their parameter cost.
  • Because the framework is designed as post-processing on a spatial-domain denoiser with matched input and output dimensions, the authors claim it can be transplanted to other spatial restoration models and high-dimensional visual data.

Reading between the lines

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

  • The gating mechanism predicts a clean ordering of gains: largest when noise is spatially non-uniform and frequency-localized (real sensor noise, stripe, mixture), smaller for i.i.d. Gaussian; a direct test would be to measure the gate's activation statistics on each noise type and confirm that the gate approaches 0 in noisy regions and 1 in clean ones.
  • Because the new modules add 61% parameters but only about 15% FLOPs, the design suggests the same frequency-gating branch could be transplanted onto other spatial-spectral backbones; if the gain transfers, the framework would be a general plug-in rather than a SERT-specific fix.
  • The ablation's single-run gaps of 0.05 to 0.15 dB between the frequency-only and baseline configurations mean the independent contribution of the frequency module is not yet separated from training noise; multi-seed runs would clarify which module actually carries the gain.
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 / 6 minor

Summary. The paper proposes HDST, a hyperspectral image denoising network built on the SERT baseline. The method adds an FFT-based preprocessing module with multi-band ASPP convolution (ASPP-FFT), a gated frequency-spatial fusion module (FSGF), a frequency-spatial collaborative attention module (FSCA), and a hybrid-domain multi-scale module (HDMS). The authors evaluate on the Realistic dataset and the ICVL synthetic dataset, reporting PSNR/SSIM/SAM comparisons against traditional and deep-learning baselines, an ablation study, and a computational efficiency analysis. The central claim is that HDST significantly improves over SERT, with +0.94 dB PSNR on Realistic and up to +0.52 dB on ICVL, while increasing FLOPs by only 14.7%.

Significance. If the reported gains are reproducible, the architecture is a plausible incremental extension of SERT with a sensible frequency-domain branch, and the released code would be a useful resource for the HSI denoising community. The paper also gives a rare explicit attempt to connect multiscale atrous convolution with frequency-band processing. However, the significance of the empirical claim is currently limited by the lack of multi-seed evaluation, by SSIM regressions on several noise types, and by the authors' own statements that figures and ablations are unfinished. The contribution is therefore better viewed as a promising system study than as a settled methodological result.

major comments (4)
  1. [Section IV-A, Table I; Section IV-B, Table II; Section IV-C, Table III] All comparisons rest on a single training run per configuration. The headline deltas are small: 0.94 dB on Realistic, 0.15-0.52 dB on ICVL, and only 0.05 dB for the frequency module alone in the ablation (Net1 vs Baseline, Table III). At these magnitudes, run-to-run variance is a plausible alternative explanation for the reported improvement. The authors should report mean and standard deviation over at least three to five seeds and, ideally, paired statistical tests over the same test images.
  2. [Table II] HDST's SSIM is lower than SERT's on Gaussian+Deadline (0.9966 vs 0.9969), Gaussian+Stripe (0.9963 vs 0.9969), and Non-i.i.d Gaussian (0.9970 vs 0.9971). The claim that HDST 'significantly outperforms' the baseline is therefore not consistent across metrics. This should be discussed explicitly, and the authors should determine whether the PSNR gains come at the cost of structural fidelity in selected cases.
  3. [Figures 3-4 and Acknowledgments] The captions state 'We are working on this figure, since we want to provide a more level of detail comparison for the visual comparison,' and the acknowledgments state 'This preprint reports work-in-progress. Full ablation studies and refined technical descriptions will be updated in subsequent versions.' These statements are incompatible with a journal submission. The visual evidence must be finalized and the work-in-progress language removed before the manuscript can be evaluated as a complete research article.
  4. [Section III.B.2 and Equations (1)-(5)] The physical-interpretation claim that ASPP dilation rates [2,4,8] correspond to multi-band frequency processing is asserted but not demonstrated. The authors should provide evidence, for example, by visualizing the learned gating masks or frequency responses and showing that they align with the stated noise priors (e.g., stripe-noise frequency lines vs high-frequency sensor noise). Without such evidence, the 'frequency domain noise prior' rationale remains post hoc.
minor comments (6)
  1. [Section III.B.2, Eq. (5)] The hyperparameter alpha controlling frequency-domain injection intensity is never given a value or sensitivity analysis; please specify its value and any effect on performance.
  2. [Section III.B.3, Eq. (6)] The FSCA window size M is not specified in the paper; please report the value used and justify it.
  3. [Table III] The column header 'ASPP/HDMS' is ambiguous for Net3 and Net4; clearly define which modules (ASPP-FFT, HDMS, dynamic fusion) are active in each ablation row.
  4. [Related Work, reference [16]] Reference [16] is described as a cross-domain filtering denoising method (FCENet), but the cited paper title is about arbitrary-shaped text detection; either the intended reference is different or the citation is incorrect.
  5. [Section IV.C] The phrase 'slight increase in parameters such as PSNR' should be reworded to 'slight increase in metrics such as PSNR' for clarity.
  6. [Title and Abstract] There are formatting glitches such as 'T ransformer' in the title line and a duplicated sentence in the abstract; please proofread the final version.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: HDST is an empirical architecture comparison with external benchmarks and no self-citation load-bearing steps.

full rationale

The paper makes no theoretical derivation claim that could reduce to its own inputs. The central evidence is empirical: HDST is compared against external benchmarks (BM4D, NGMeet, HSID-CNN, MAC-Net, SSRT-UNet) and a baseline SERT [20] on fixed real and synthetic splits, with the paper stating 'We followed the SERT division method, using the same 44 images and 15 images for training and testing.' Every named module (FFT, ASPP, attention gating) is a standard component whose citations are external prior work, not the authors' own. The only potentially load-bearing comparison is the SERT baseline, but SERT is cited from other authors (M. Li et al., arXiv:2304.00844), not self-citation. The paper's claims about frequency-domain noise priors are explanatory commentary on observed results, not used as a premise to derive the numerical outcomes. The single-run evaluation and lack of significance tests are methodological robustness concerns, not circularity. Therefore no circular step meets the evidentiary standard of Eq. X = Eq. Y by construction or fitted-parameter-renamed-as-prediction.

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

The central claim depends on no new physical entities or fitted constants beyond the network's learned parameters and a few hand-chosen hyperparameters. The main axioms are domain assumptions about noise separability and synthetic-to-real transfer, which are plausible but not rigorously established.

free parameters (4)
  • ASPP-FFT dilation rates = [2,4,8]
    Hand-selected to represent multi-band frequency decomposition; the claim that these rates map to frequency bands is heuristic, and no sensitivity study is reported.
  • Dynamic fusion coefficient beta = 0.1 (initial)
    Learnable but initialized to 0.1; the residual injection strength is partly chosen by hand.
  • Gating injection intensity alpha
    Hyperparameter controlling frequency-domain information injection; no value or sensitivity analysis is given.
  • FSCA window size M
    Window division for attention is not quantified in the paper, but affects computational cost and receptive field.
assumptions (4)
  • standard math FFT/IFFT are invertible and preserve information
    Relied on in Eq. (1)-(3); standard.
  • domain assumption Hyperspectral noise is separable in the frequency domain and concentrated in high frequencies
    Central motivation in Section III.B: the paper assumes real HSI noise energy is concentrated in high-frequency bands, which justifies the frequency post-processing module. This is an empirical assumption, not proven.
  • domain assumption Synthetic noise patterns (Non-i.i.d Gaussian, stripe, deadline, impulse) approximate real sensor noise
    Used in Section IV.B to generate test data; the paper itself notes discrepancies between synthetic and real noise, so the transferability is questionable.
  • domain assumption The gating mechanism can reliably distinguish clean and noisy regions via the learned mask
    Eq. (4)-(5): Gate is a learned mask that the paper asserts preserves clean areas and replaces noisy regions; this capacity is not theoretically guaranteed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hybrid-Domain Synergistic Transformer for Hyperspectral Image Denoising." pith.science (2026). https://pith.science/paper/24KYAUCH

@misc{pith2026250720099,
  author       = {Pith},
  title        = {Pith review of: Hybrid-Domain Synergistic Transformer for Hyperspectral Image Denoising},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/24KYAUCH}},
  note         = {Machine review of arXiv:2507.20099}
}
read the original abstract

Hyperspectral image denoising faces the challenge of multi-dimensional coupling of spatially non-uniform noise and spectral correlation interference. Existing deep learning methods mostly focus on RGB images and struggle to effectively handle the unique spatial-spectral characteristics and complex noise distributions of hyperspectral images (HSI). This paper proposes an HSI denoising framework, Hybrid-Domain Synergistic Transformer Network (HDST), based on frequency domain enhancement and multiscale modeling, achieving three-dimensional collaborative processing of spatial, frequency and channel domains. The method innovatively integrates three key mechanisms: (1) introducing an FFT preprocessing module with multi-band convolution to extract cross-band correlations and decouple spectral noise components; (2) designing a dynamic cross-domain attention module that adaptively fuses spatial domain texture features and frequency domain noise priors through a learnable gating mechanism; (3) building a hierarchical architecture where shallow layers capture global noise statistics using multiscale atrous convolution, and deep layers achieve detail recovery through frequency domain postprocessing. Experiments on both real and synthetic datasets demonstrate that HDST significantly improves denoising performance while maintaining computational efficiency, validating the effectiveness of the proposed method. This research provides new insights and a universal framework for addressing complex noise coupling issues in HSI and other high-dimensional visual data. The code is available at https://github.com/lhy-cn/HDST-HSIDenoise.

Figures

Figures reproduced from arXiv: 2507.20099 by the authors.

Figure 1
Figure 1. FIG. 1. Overall framework of HDST [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Cross-domain feedback loop [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Visualization of denoising results on Realistic dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Visualization of denoising results on ICVL synthetic dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 33 canonical work pages

  1. [1]

    merlin.mbs aapmrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aapmrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translat...

    2010

  2. [2]

    merlin.mbs aipauth4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aipauth4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translat...

    2010

  3. [3]

    merlin.mbs aipnum4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aipnum4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translati...

    2010

  4. [4]

    merlin.mbs apsrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs apsrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translati...

    2010

  5. [5]

    merlin.mbs apsrmp4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs apsrmp4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translati...

    2010

  6. [6]

    Yang , author C

    author author C. Yang , author C. Zhang , author H. Shen , author T. Peng , author C. Wang , author L. Deng , author H. Chen , \ and\ author L. He ,\ title title Hfan: High-frequency attention network for hyperspectral image denoising , \ 10.1007/s13042-023-01942-2 journal journal International Journal of Machine Learning and Cybernetics \ volume 15 ,\ pa...

    work page doi:10.1007/s13042-023-01942-2

  7. [7]

    Lin , author Z

    author author Y. Lin , author Z. Cai , author J. Li , \ and\ author J. Zhang ,\ title title Lightweight remote sensing image denoising via knowledge distillation , \ in\ 10.1109/MMSP55362.2022.9949236 booktitle 2022 IEEE 24th International Workshop on Multimedia Signal Processing (MMSP) \ ( year 2022 )\ pp.\ pages 1--7 NoStop

    arXiv 2022

  8. [9]

    Li \ and\ author F

    author author Y. Li \ and\ author F. Zhou ,\ title title Research on denoising optimization algorithm of point cloud in 3d reconstruction , \ https://api.semanticscholar.org/CorpusID:213059298 journal journal Journal of Physics: Conference Series \ volume 1345 ( year 2019 ) NoStop

    work page 2019

Show all 49 references
  1. [10]

    Rasti , author J

    author author B. Rasti , author J. R. \ Sveinsson , author M. O. \ Ulfarsson , \ and\ author J. A. \ Benediktsson ,\ title title Hyperspectral image denoising using 3d wavelets , \ in\ 10.1109/IGARSS.2012.6351286 booktitle 2012 IEEE International Geoscience and Remote Sensing ...

    2012

  2. [11]

    Kaur \ and\ author Neetu ,\ title title Comparison of non local means filtering with wavelet transform technique for denoising ecg signal , \ \ ( year 2015 ) NoStop

    author author N. Kaur \ and\ author Neetu ,\ title title Comparison of non local means filtering with wavelet transform technique for denoising ecg signal , \ \ ( year 2015 ) NoStop

    2015

  3. [12]

    Zhao , author Y

    author author Y. Zhao , author Y. Li , author X. Dong , \ and\ author B. Yang ,\ title title Low-frequency noise suppression method based on improved dncnn in desert seismic data , \ @noop journal journal IEEE Geoscience and Remote Sensing Letters \ volume 16 ,\ pages 811--815...

    2018

  4. [13]

    Liang , author J

    author author J. Liang , author J. Cao , author G. Sun , author K. Zhang , author L. Van Gool , \ and\ author R. Timofte ,\ title title Swinir: Image restoration using swin transformer , \ in\ @noop booktitle Proceedings of the IEEE/CVF international conference on computer vis...

    2021

  5. [14]

    author author M. J. \ Khan , author H. S. \ Khan , author A. Yousaf , author K. Khurshid , \ and\ author A. Abbas ,\ title title Modern trends in hyperspectral image analysis: A review , \ 10.1109/ACCESS.2018.2812999 journal journal IEEE Access \ volume 6 ,\ pages 14118--14129...

    2018

  6. [15]

    Maffei , author J

    author author A. Maffei , author J. M. \ Haut , author M. E. \ Paoletti , author J. Plaza , author L. Bruzzone , \ and\ author A. Plaza ,\ title title A single model cnn for hyperspectral image denoising , \ 10.1109/TGRS.2019.2952062 journal journal IEEE Transactions on Geosci...

  7. [16]

    Xiong , author J

    author author F. Xiong , author J. Zhou , author S. Tao , author J. Lu , author J. Zhou , \ and\ author Y. Qian ,\ title title Smds-net: Model guided spectral-spatial network for hyperspectral image denoising , \ 10.1109/TIP.2022.3196826 journal journal IEEE Transactions on Im...

  8. [17]

    Ozdemir \ and\ author K

    author author A. Ozdemir \ and\ author K. Polat ,\ title title Deep learning applications for hyperspectral imaging: A systematic review , \ 10.33969/JIEC.2020.21004 journal journal Journal of the Institute of Electronics and Computer \ volume 2 ,\ pages 39--56 ( year 2020 ) NoStop

    2020

  9. [18]

    Zhang , author H

    author author H. Zhang , author H. Xu , author X. Yu , author X. Zhang , \ and\ author C. Wu ,\ title title Leveraging frequency and spatial domain information for underwater image restoration , \ 10.1088/1742-6596/2832/1/012001 journal journal Journal of Physics: Conference S...

    doi

  10. [19]

    Zheng , author K

    author author P. Zheng , author K. Chen , author J. Huang , author B. Chen , author J. Liu , author Y. Ren , \ and\ author X. Pu ,\ https://arxiv.org/abs/2504.11286 title Lightweight medical image restoration via integrating reliable lesion-semantic driven prior , \ ( year 202...

    2025 arXiv

  11. [20]

    Gao , author X

    author author N. Gao , author X. Jiang , author X. Zhang , \ and\ author Y. Deng ,\ title title Efficient frequency-domain image deraining with contrastive regularization , \ in\ @noop booktitle Computer Vision -- ECCV 2024 ,\ editor edited by\ editor A. Leonardis , editor E. ...

    2024

  12. [21]

    Zhu , author J

    author author Y. Zhu , author J. Chen , author L. Liang , author Z. Kuang , author L. Jin , \ and\ author W. Zhang ,\ title title Fourier contour embedding for arbitrary-shaped text detection , \ in\ @noop booktitle Proceedings of the IEEE/CVF Conference on Computer Vision and...

    2021

  13. [22]

    Chen , author G

    author author L. Chen , author G. Papandreou , author I. Kokkinos , author K. Murphy , \ and\ author A. L. \ Yuille ,\ title title Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs , \ http://arxiv.org/abs/1606.0091...

    2016 arXiv

  14. [23]

    Lian , author Y

    author author X. Lian , author Y. Pang , author J. Han , \ and\ author J. Pan ,\ title title Cascaded hierarchical atrous spatial pyramid pooling module for semantic segmentation , \ https://doi.org/10.1016/j.patcog.2020.107622 journal journal Pattern Recognition \ volume 110 ...

    2020

  15. [24]

    Zhu , author L

    author author Y. Zhu , author L. Wan , author W. Xu , \ and\ author S. Wang ,\ title title Aspp-df-pvnet: Atrous spatial pyramid pooling and distance-filtered pvnet for occlusion resistant 6d object pose estimation , \ https://doi.org/10.1016/j.image.2021.116268 journal journa...

  16. [25]

    Li , author J

    author author M. Li , author J. Liu , author Y. Fu , author Y. Zhang , \ and\ author D. Dou ,\ https://arxiv.org/abs/2304.00844 title Spectral enhanced rectangle transformer for hyperspectral image denoising , \ ( year 2023 ),\ http://arxiv.org/abs/2304.00844 arXiv:2304.00844 ...

    2023 arXiv

  17. [26]

    Palsson , author M

    author author F. Palsson , author M. O. \ Ulfarsson , \ and\ author J. R. \ Sveinsson ,\ title title Hyperspectral image denoising using a sparse low rank model and dual-tree complex wavelet transform , \ https://api.semanticscholar.org/CorpusID:2098232 journal journal 2014 IE...

    2014

  18. [27]

    Rasti , author P

    author author B. Rasti , author P. Scheunders , author P. Ghamisi , author G. Licciardi , \ and\ author J. Chanussot ,\ title title Noise reduction in hyperspectral imagery: Overview and application , \ @noop journal journal Remote Sensing \ volume 10 ,\ pages 482 ( year 2018 ) NoStop

    2018

  19. [28]

    Zhou , author Q.-J

    author author D. Zhou , author Q.-J. \ Wang , author Q.-J. \ Tian , author Q.-Z. \ Lin , author W.-X. \ Fu , et al. ,\ title title Wavelet analysis and its application in denoising the spectrum of hyperspectral image , \ @noop journal journal Spectroscopy and Spectral Analysis...

    1941

  20. [29]

    Xie \ and\ author Y

    author author W. Xie \ and\ author Y. Li ,\ title title Hyperspectral imagery denoising by deep learning with trainable nonlinearity function , \ 10.1109/LGRS.2017.2743738 journal journal IEEE Geoscience and Remote Sensing Letters \ volume 14 ,\ pages 1963--1967 ( year 2017 ) NoStop

    2017

  21. [30]

    Dong , author H

    author author W. Dong , author H. Wang , author F. Wu , author G. Shi , \ and\ author X. Li ,\ title title Deep spatial–spectral representation learning for hyperspectral image denoising , \ https://api.semanticscholar.org/CorpusID:146116742 journal journal IEEE Transactions o...

    2019

  22. [31]

    Liang , author C

    author author H. Liang , author C. Ke , \ and\ author K. Li ,\ title title Hybrid spatial-spectral neural network for hyperspectral image denoising , \ in\ @noop booktitle Computer Vision -- ECCV 2024 Workshops ,\ editor edited by\ editor A. Del Bue , editor C. Canton , editor...

    2024

  23. [32]

    Fu , author F

    author author G. Fu , author F. Xiong , author S. Tao , author J. Lu , author J. Zhou , \ and\ author Y. Qian ,\ title title Learning a model-based deep hyperspectral denoiser from a single noisy hyperspectral image , \ in\ 10.1109/IGARSS47720.2021.9553257 booktitle 2021 IEEE ...

  24. [33]

    Gao , author X

    author author N. Gao , author X. Jiang , author X. Zhang , \ and\ author Y. Deng ,\ title title Efficient frequency-domain image deraining with contrastive regularization , \ in\ @noop booktitle European Conference on Computer Vision \ ( organization Springer ,\ year 2024 )\ p...

    2024

  25. [34]

    Kong , author J

    author author L. Kong , author J. Dong , author J. Ge , author M. Li , \ and\ author J. Pan ,\ title title Efficient frequency domain-based transformers for high-quality image deblurring , \ in\ @noop booktitle Proceedings of the IEEE/CVF Conference on Computer Vision and Patt...

    2023

  26. [35]

    Wang , author H

    author author Y. Wang , author H. Wang , author L. Wang , author X. Wang , author L. Zhu , author W. Lu , \ and\ author H. Huang ,\ title title Complementary advantages: Exploiting cross-field frequency correlation for nir-assisted image denoising , \ in\ @noop booktitle Proce...

    2025

  27. [36]

    Wang , author L

    author author S. Wang , author L. Li , author X. Li , author J. Zhang , author L. Zhao , author X. Su , \ and\ author F. Chen ,\ title title A denoising network based on frequency-spectral- spatial-feature for hyperspectral image , \ 10.1109/JSTARS.2023.3285454 journal journal...

  28. [37]

    Sheng , author Z

    author author Z. Sheng , author Z. Yu , author X. Liu , author S.-Y. \ Cao , author Y. Liu , author H.-L. \ Shen , \ and\ author H. Zhang ,\ title title Structure aggregation for cross-spectral stereo image guided denoising , \ in\ @noop booktitle Proceedings of the IEEE/CVF C...

    2023

  29. [38]

    Zhang , author Y

    author author Q. Zhang , author Y. Zheng , author Q. Yuan , author M. Song , author H. Yu , \ and\ author Y. Xiao ,\ title title Hyperspectral image denoising: From model-driven, data-driven, to model-data-driven , \ 10.1109/TNNLS.2023.3278866 journal journal IEEE Transactions...

  30. [39]

    Archana \ and\ author P

    author author R. Archana \ and\ author P. S. E. \ Jeevaraj ,\ title title Deep learning models for digital image processing: a review , \ 10.1007/s10462-023-10631-z journal journal Artificial Intelligence Review \ volume 57 ,\ pages 11 ( year 2024 ) NoStop

    doi

  31. [40]

    Maggioni , author V

    author author M. Maggioni , author V. Katkovnik , author K. Egiazarian , \ and\ author A. Foi ,\ title title Nonlocal transform-domain filter for volumetric data denoising and reconstruction , \ @noop journal journal IEEE transactions on image processing \ volume 22 ,\ pages 1...

    2012

  32. [41]

    He , author Q

    author author W. He , author Q. Yao , author C. Li , author N. Yokoya , \ and\ author Q. Zhao ,\ title title Non-local meets global: An integrated paradigm for hyperspectral denoising , \ in\ @noop booktitle Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern...

    2019

  33. [42]

    author author A. Brook ,\ title title 3-d wavelets-based denoising and enhancement of hyperspectral imagery , \ in\ 10.1109/WHISPERS.2014.8077589 booktitle 2014 6th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS) \ ( year 2014 )\ p...

    2014

  34. [43]

    Yuan , author Q

    author author Q. Yuan , author Q. Zhang , author J. Li , author H. Shen , \ and\ author L. Zhang ,\ title title Hyperspectral image denoising employing a spatial–spectral deep residual convolutional neural network , \ 10.1109/TGRS.2018.2865197 journal journal IEEE Transactions...

  35. [44]

    Xiong , author J

    author author F. Xiong , author J. Zhou , author Q. Zhao , author J. Lu , \ and\ author Y. Qian ,\ title title Mac-net: Model-aided nonlocal neural network for hyperspectral image denoising , \ @noop journal journal IEEE Transactions on Geoscience and Remote Sensing \ volume 6...

    2021

  36. [45]

    Fu , author F

    author author G. Fu , author F. Xiong , author J. Lu , author J. Zhou , author J. Zhou , \ and\ author Y. Qian ,\ title title Hyperspectral image denoising via spatial–spectral recurrent transformer , \ 10.1109/TGRS.2024.3374953 journal journal IEEE Transactions on Geoscience ...

    2024

  37. [46]

    Sara , author M

    author author U. Sara , author M. Akter , \ and\ author M. S. \ Uddin ,\ title title Image quality assessment through fsim, ssim, mse and psnr—a comparative study , \ @noop journal journal Journal of Computer and Communications \ volume 7 ,\ pages 8--18 ( year 2019 ) NoStop

    2019

  38. [47]

    Terven , author D.-M

    author author J. Terven , author D.-M. \ Cordova-Esparza , author J.-A. \ Romero-Gonz \'a lez , author A. Ram \' rez-Pedraza , \ and\ author E. Ch \'a vez-Urbiola ,\ title title A comprehensive survey of loss functions and metrics in deep learning , \ @noop journal journal Art...

    2025

  39. [48]

    Zhang , author L

    author author H. Zhang , author L. Liu , author W. He , \ and\ author L. Zhang ,\ title title Hyperspectral image denoising with total variation regularization and nonlocal low-rank tensor decomposition , \ 10.1109/TGRS.2019.2947333 journal journal IEEE Transactions on Geoscie...

    2019

  40. [49]

    Wei , author Y

    author author K. Wei , author Y. Fu , \ and\ author H. Huang ,\ title title 3-D Quasi-Recurrent Neural Network for Hyperspectral Image Denoising , \ 10.1109/TNNLS.2020.2978756 journal journal IEEE Transactions on Neural Networks and Learning Systems \ volume 32 ,\ pages 363--3...

    2020

  41. [50]

    Bodrito , author A

    author author T. Bodrito , author A. Zouaoui , author J. Chanussot , \ and\ author J. Mairal ,\ https://arxiv.org/abs/2111.09708 title A trainable spectral-spatial sparse coding model for hyperspectral image restoration , \ ( year 2021 ),\ http://arxiv.org/abs/2111.09708 arXiv...

    2021

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.