Pith. sign in

REVIEW 4 major objections 5 minor 58 references

Attentive Contextual Attention for Cloud Removal

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper claims that softmax attention weights are never exactly zero, so cloud-removal outputs inherit noise from irrelevant cloud-covered patches, and that AC-Attention fixes this by learning a per-query rescaling that sets useless…

desk verdict AC-Attention is a small but plausible tweak to Contextual Attention with decent evidence on RICE, but the SEN12MS-CR claim rests on uncontrolled imported baselines. read the letter →

arxiv 2411.13042 v1 pith:NEHPKMMB submitted 2024-11-20 cs.CV eess.IV

Wenli Huang , Ye Deng , Yang Wu , Jinjun Wang This is my paper · ORCID
classification cs.CVeess.IV
keywords cloudremovalattentionmechanismremotesensingimageryimagerestorationAttentiveContextualACA-CRNetDSen2-CRSEN12MS-CR
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

Clouds cover much of the Earth's surface in remote sensing imagery, and the deep learning models used to remove them can return blurry, inconsistent images. This paper argues that one culprit is the attention mechanism itself: softmax-based similarity scores are never exactly zero, so every output patch is a weighted average that includes noise from cloud-covered or unrelated patches. The proposed answer, Attentive Contextual Attention (AC-Attention), learns a per-query weight and bias that rescale the similarity matrix and then applies ReLU, zeroing out irrelevant connections before they reach the output. Built into the DSen2-CR residual network as ACA-CRNet, the module reports better reconstruction on all four metrics across RICE-I, RICE-II, and SEN12MS-CR than six comparison methods. The same module also improves existing networks when dropped into them, so the paper's claim is not just a single tuned architecture but a reusable correction to how attention is computed.

What carries the argument

The load-bearing object is the AC-Attention module, a patch-based attention block. After the input feature $F$ is projected to query, key, and value embeddings and split into $s\times s$ patches, it computes the patch similarity matrix $S_p = \mathrm{softmax}(Q_p K_p^T / \sqrt{d})$. Two small convolutional modules read the query $Q$ and emit a weight $W$ and bias $B$; the attentive score matrix is $S_{att} = \mathrm{ReLU}[(S_p - \mathrm{Avg}(S_p)) W + B]$, and the output is $O_p = S_{att} V_p$, where $\mathrm{Avg}$ is the row-wise average. This transformation is what carries the argument: it lets the network learn, for each query, which long-range matches to keep and which cloud-corrupted or irrelevant matches to set to zero.

What would settle it

Re-run all six baselines on the same SEN12MS-CR split, preprocessing, crop size, and evaluation script used for ACA-CRNet, without importing numbers from another paper; if any baseline then matches or beats ACA-CRNet on MAE, SAM, PSNR, or SSIM, the claimed consistent superiority on that dataset is not established.

Watch

Extended reading notes

Core claim

The central claim is that the failure mode of global attention in cloud removal is not a small receptive field but an indiscriminate weighting of every patch. For a query patch $q_i$, the vanilla softmax similarity $s_{i,j}$ is positive for all key patches $j$, so patches with no real relationship—often cloud-covered regions—still contribute to the reconstruction and blur the result. AC-Attention replaces this with $S_{att} = \mathrm{ReLU}[(S_p - \mathrm{Avg}(S_p)) W + B]$, where the row-average subtraction centers the similarity matrix and the query-dependent $W$ and $B$ rescale it, so negative and near-zero relations are discarded. The paper embeds this mechanism in two Residual AC-Attention Blocks inside the DSen2-CR residual network, yielding ACA-CRNet; on RICE-I, RICE-II, and SEN12MS-CR it reports the best MAE, SAM, PSNR, and SSIM among the compared methods, and ablations show similar gains when AC-Attention replaces ordinary blocks in SpA GAN, STGAN, encoder-decoder, and U-shaped architectures.

Load-bearing premise

The SEN12MS-CR comparison assumes that metric values taken from the UnCRtainTS paper for SpA GAN, DSen2-CR, GLF-CR, and UnCRtainTS are directly comparable to ACA-CRNet numbers produced under the authors' own cropping, batch size, and epoch settings, even though those baselines were not retrained under the same protocol.

Editorial extensions

If this is right

  • ACA-CRNet reports the best scores on all four metrics on RICE-I, RICE-II, and SEN12MS-CR among the six compared methods; the reported PSNR values are 36.66, 33.70, and 29.78 dB on the three datasets.
  • Swapping AC-Attention into existing cloud-removal models (SpA GAN, STGAN, and DSen2-CR) improves their MAE, SAM, PSNR, and SSIM by substantial average margins, so the module is a drop-in component for those architectures, not only for the authors' own network.
  • AC-Attention also improves encoder-decoder and U-shaped networks compared with both their base versions and versions using Contextual Attention, supporting the claim that the selection step, not the surrounding architecture, drives the gain.
  • The similarity-score visualizations show AC-Attention concentrates high scores on semantically matching ground regions and assigns zero to cloud-covered areas, meaning the mechanism's behavior can be inspected directly in a trained model.

Reading between the lines

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

  • Beyond cloud removal, the same softmax-never-zero problem appears in image inpainting and dehazing, where distant irrelevant patches also get small nonzero weights; a straightforward extension is to install AC-Attention in those tasks and run the same base-versus-CA-versus-AC ablation protocol.
  • Because $W$ and $B$ are produced solely from the query, the module needs no temporal or SAR inputs; I infer it could be combined with temporal-SAR fusion models to get both long-range context and cloud-patch filtering, a configuration the paper does not test.
  • The paper's visualization suggests a quantitative check: for queries located inside cloud-covered regions, the learned row sums of $S_{att}$ over cloud patches should be near zero while rows for clear-sky queries concentrate on relevant patches; measuring this across SEN12MS-CR would test whether the mechanism behaves as described.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes Attentive Contextual Attention (AC-Attention), a patch-based attention mechanism that adds a learned per-query linear transformation to the similarity scores of contextual attention, with the stated goal of suppressing irrelevant long-range dependencies in cloud removal. The authors embed AC-Attention into the DSen2-CR residual architecture to form ACA-CRNet, replacing two residual blocks with residual AC-attention blocks. They report quantitative results on RICE-I, RICE-II, and SEN12MS-CR, claiming that ACA-CRNet outperforms six baselines on all four metrics on all three datasets, and they provide ablations in which AC-Attention is inserted into three existing methods and three network architectures. The paper also presents visualizations of attention similarity maps and cloud-free outputs.

Significance. If the central claim holds, the contribution is a practical and reasonably simple module: a learned gating/bias mechanism on attention scores that improves reconstruction quality in cloud removal, with demonstrated gains over several published baselines. The strengths of the paper include evaluation on public benchmarks, use of standard metrics, availability of code, and an ablation design that tests the module across multiple architectures. However, the significance is currently tempered by an uncontrolled comparison on the SEN12MS-CR dataset, a lack of error bars or significance tests, and several numerical inconsistencies in the reported averages and ablation tables. The method itself is not circular: the central mechanism is a learned module trained on the target data, and the evaluation uses external benchmarks.

major comments (4)
  1. [Table III and Section IV.B.1] The SEN12MS-CR comparison is not controlled. The table footnote states that the metric values for SpA GAN, DSen2-CR, GLF-CR, and UnCRtainTS are sourced from the UnCRtainTS paper, whereas ACA-CRNet is trained under the authors' own protocol described in Section IV.A.3 (256x256 crops, batch size 6, 30 epochs). The paper does not establish that the imported values use the same test split, input scaling, preprocessing, or metric implementation as the ACA-CRNet evaluation. Because the SEN12MS-CR margins are small (MAE 0.025 vs. 0.027, PSNR 29.78 vs. 28.90, SSIM 0.896 vs. 0.880), a modest protocol mismatch could change the ranking. This is the load-bearing evidence for the claim that ACA-CRNet 'consistently outperforms all baseline methods' on all three datasets, so the authors should either retrain the baselines under their own protocol or obtain and report official numbers produced under a matched evaluation protocol, for the SEN12MS-CR rows at minimum.
  2. [Section IV.B.2] The claimed average improvements over UnCRtainTS do not match the numbers in Table III. For MAE, the per-dataset relative improvements are approximately 36.4%, 20.0%, and 7.4%, whose arithmetic mean is about 21.3%, not the reported 30.05%. For SAM, the corresponding mean is approximately 16.6%, not 21.05%; for PSNR it is approximately 6.4%, not 5.95%. The SSIM average is close to the reported 1.58%, but the others are not. The authors should either correct the averages or state the exact formula used to aggregate them.
  3. [Table IV] The ablation table contains a clearly anomalous entry: on RICE-II, ACA-EDNet with AC-Attention reports MAE = 0.2093, whereas the base encoder-decoder reports 0.0218 and the CA version reports 0.021. This single entry contradicts the paper's statement that replacing CA with AC-Attention leads to consistent improvements across all architectures, and it is almost certainly a typo or an evaluation error. The authors should verify the entry and rerun or report the corrected value, and they should re-examine the surrounding numbers for similar transcription errors.
  4. [Section III.A.2, Eq. (6)] As written, the operation in Eq. (6) is not query-dependent in the way the text claims. Because each row of the softmax matrix Sp sums to 1, the row average Sp,avg is identically 1/Np for every query. Subtracting Sp,avg therefore subtracts a constant shift from every row of Sp, so the 'adjusted similarity matrix' Sp,ad does not capture query-specific average similarity. The query-dependent behavior must come entirely from the learned W and B terms in Eq. (7). The authors should either correct the description of Eq. (6) or clarify the intended computation, and they should also specify how the pixel-level W and B tensors from Eq. (5) are broadcast or reshaped to match the patch-level matrix Sp.
minor comments (5)
  1. [Section II.B] The reference to 'Evel et al.' for the UnCRtainTS method should be 'Ebel et al.' to match the cited author in [10].
  2. [Section IV.A.3] The implementation details state that the base channel size C is 256, but Table I shows that the input channel count Cin varies with the dataset; the paper should clarify how the multispectral 13-band SEN12MS-CR data are mapped to the network's input channels.
  3. [Table V] In Table V, Ours-STGAN on RICE-I shows a SAM increase from 1.701 to 2.223 relative to the base STGAN, while the text says the modifications produced substantial improvements in all metrics; this exception should be acknowledged or explained.
  4. [Section IV.B.2] The phrase 'significant improvements' is used for the quantitative gains, but no statistical significance tests or error bars are reported; the authors should soften the wording or provide variance estimates across multiple runs or test sets.
  5. [Section III.A.2] The notation in Fig. 2 is dense and partly illegible, particularly the expressions for the patch counts and reshaped dimensions; a cleaner definition of the patch index mapping would improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: AC-Attention is a learned module evaluated on external benchmarks; no equation or fitted parameter reduces to the conclusion, and the only self-citation [56] is a non-load-bearing architecture choice.

full rationale

The paper's derivation chain contains no step that equates a prediction to its own input. AC-Attention is defined by explicit operations (Eqs. 4-8): S_p = softmax(Q_p K_p^T / sqrt(d)), S_att = ReLU[(S_p - S_p,avg)*W + B], and O_p = S_att V_p, with W and B learned from Q via small conv-ReLU modules. These are architectural components trained with a reconstruction loss, not parameters fitted to the evaluation metric and then renamed as predictions. The central claim that ACA-CRNet outperforms baselines is an empirical result on public benchmarks (RICE-I, RICE-II, SEN12MS-CR) and could plausibly have gone the other way; no equation in the paper forces it. The only self-citation is reference [56], used as the encoder-decoder host network (ACA-EDNet) in an ablation study, not as evidence for the headline performance claim, so it is not load-bearing. Two numerical/comparability concerns should not be mistaken for circularity: the Table III footnote says metrics for methods marked * are 'sourced from the UnCRtainTS paper [10]', which raises protocol-matching questions for SEN12MS-CR baselines, and the reported 'average gains of 30.05% and 21.05%' do not match a direct recomputation from Table III (MAE average gain vs UnCRtainTS is about 21%, not 30%). These are correctness and external-validity issues, not cases where the paper's conclusion is identical to its assumptions by construction. The method is also evaluated against external baselines and public datasets, so the self-contained benchmark evidence keeps circularity at zero.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper makes an empirical performance claim, so the ledger lists hand-chosen architecture and training choices plus dataset assumptions rather than derivation constants. The network's learned weights are the method itself, not hidden fit parameters. The most consequential assumptions are the comparability of imported baseline metrics from UnCRtainTS and the validity of the dataset references as ground truth.

free parameters (4)
  • Base channel size C = 256
    Width of all feature maps in ACA-CRNet, chosen to align with the DSen2-CR baseline; stated in Section IV.A.3 and Table I.
  • Residual scaling factor alpha = 0.1
    Fixed scale for residual branches in RB and RACAB, inherited from DSen2-CR; stated in Section III.B.2 and Table I.
  • AC-Attention patch size s = not specified
    The module divides features into s x s patches (Section III.A.2 and Fig. 2), but the value of s is never stated, which is a reproducibility gap.
  • Training schedule = lr=7e-5, batch 12 or 6, epochs 300 or 30
    Hyperparameters listed in Section IV.A.3 are central to reproducing the reported metrics, but they are not justified by sensitivity analysis.
assumptions (5)
  • standard math Standard softmax attention and patch-based matching formulas from Vaswani et al. and Yu et al. are used without proof.
    Section III.A.1 and III.A.2 rely on the standard scaled dot-product attention and Contextual Attention patch matching as background.
  • domain assumption The cloud-free references in RICE-I, RICE-II, and SEN12MS-CR are treated as ground truth for reconstruction quality.
    All four metrics compare predictions against these references; any residual cloud, haze, or misregistration in the references would bias scores. See dataset descriptions in Section IV.A.1.
  • domain assumption The SEN12MS-CR baseline numbers imported from UnCRtainTS [10] are directly comparable to the authors' own training results.
    Table III footnote says starred results are sourced from [10] while Ours is trained under the authors' protocol; comparability of splits, preprocessing, and evaluation is assumed.
  • ad hoc to paper A per-query linear transformation learned by two stacked convolutions can learn to suppress irrelevant attention scores in cloud removal.
    This is the core design hypothesis of AC-Attention, supported only by empirical ablations in Section IV.C, with no theoretical justification.
  • domain assumption The DSen2-CR residual architecture, reconstruction loss, and residual scaling alpha=0.1 remain appropriate after inserting AC-Attention blocks.
    The network is assembled by swapping two residual blocks into DSen2-CR; the paper does not analyze whether the original loss and scaling are still optimal. See Section III.B.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Attentive Contextual Attention for Cloud Removal." pith.science (2026). https://pith.science/paper/NEHPKMMB

@misc{pith2026241113042,
  author       = {Pith},
  title        = {Pith review of: Attentive Contextual Attention for Cloud Removal},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NEHPKMMB}},
  note         = {Machine review of arXiv:2411.13042}
}
read the original abstract

Cloud cover can significantly hinder the use of remote sensing images for Earth observation, prompting urgent advancements in cloud removal technology. Recently, deep learning strategies have shown strong potential in restoring cloud-obscured areas. These methods utilize convolution to extract intricate local features and attention mechanisms to gather long-range information, improving the overall comprehension of the scene. However, a common drawback of these approaches is that the resulting images often suffer from blurriness, artifacts, and inconsistencies. This is partly because attention mechanisms apply weights to all features based on generalized similarity scores, which can inadvertently introduce noise and irrelevant details from cloud-covered areas. To overcome this limitation and better capture relevant distant context, we introduce a novel approach named Attentive Contextual Attention (AC-Attention). This method enhances conventional attention mechanisms by dynamically learning data-driven attentive selection scores, enabling it to filter out noise and irrelevant features effectively. By integrating the AC-Attention module into the DSen2-CR cloud removal framework, we significantly improve the model's ability to capture essential distant information, leading to more effective cloud removal. Our extensive evaluation of various datasets shows that our method outperforms existing ones regarding image reconstruction quality. Additionally, we conducted ablation studies by integrating AC-Attention into multiple existing methods and widely used network architectures. These studies demonstrate the effectiveness and adaptability of AC-Attention and reveal its ability to focus on relevant features, thereby improving the overall performance of the networks. The code is available at \url{https://github.com/huangwenwenlili/ACA-CRNet}.

Figures

Figures reproduced from arXiv: 2411.13042 by the authors.

Figure 1
Figure 1. A comparison between the Contextual Attention (CA) and our proposed Attentive Contextual Attention (AC-Attention). The first two columns show [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The architecture of Attentive Contextual Attention (AC-Attention), consisting of three steps: embedding and reshaping, attentive matching, and [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Overview of our proposed ACA-CRNet, including (a) ACA-CRNet architecture, (b) Stem, (c) Residual Block (RB), (d) Residual AC-Attention Block [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Visualization of cloud removal results on RICE-I and RICE-II datasets. Local details are highlighted in red boxes. Zooming in is recommended for [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Visualization of cloud removal results for Sentinel-2 satellite data on the SEN12MS-CR dataset. The Sentinel-2 data includes 13 spectral bands, with [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Visualization results of ACA-CRNet demonstrating the effectiveness [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 49 canonical work pages

  1. [1]

    Spatial and temporal distribution of clouds observed by modis onboard the terra and aqua satellites,

    M. D. King, S. Platnick, W. P. Menzel, S. A. Ackerman, and P. A. Hubanks, “Spatial and temporal distribution of clouds observed by modis onboard the terra and aqua satellites,” IEEE transactions on geoscience and remote sensing , vol. 51, no. 7, pp. 3826–3852, 2013

    work page 2013

  2. [2]

    Simultaneous cloud detection and removal from bitemporal remote sensing images using cascade convolutional neural networks,

    S. Ji, P. Dai, M. Lu, and Y . Zhang, “Simultaneous cloud detection and removal from bitemporal remote sensing images using cascade convolutional neural networks,” IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 1, pp. 732–748, 2020

    work page 2020

  3. [3]

    Cloud removal in sentinel-2 imagery using a deep residual neural network and sar-optical data fusion,

    A. Meraner, P. Ebel, X. X. Zhu, and M. Schmitt, “Cloud removal in sentinel-2 imagery using a deep residual neural network and sar-optical data fusion,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 166, pp. 333–346, 2020

    work page 2020

  4. [4]

    Sen12ms-cr-ts: A remote- sensing data set for multimodal multitemporal cloud removal,

    P. Ebel, Y . Xu, M. Schmitt, and X. X. Zhu, “Sen12ms-cr-ts: A remote- sensing data set for multimodal multitemporal cloud removal,” IEEE Transactions on Geoscience and Remote Sensing , vol. 60, pp. 1–14, 2022

    work page 2022

  5. [5]

    Multisensor data fusion for cloud removal in global and all-season sentinel-2 imagery,

    P. Ebel, A. Meraner, M. Schmitt, and X. X. Zhu, “Multisensor data fusion for cloud removal in global and all-season sentinel-2 imagery,” IEEE Transactions on Geoscience and Remote Sensing , vol. 59, no. 7, pp. 5866–5878, 2020

    work page 2020

  6. [6]

    Image-to-image translation with conditional adversarial networks,

    P. Isola, J.-Y . Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 1125– 1134

    2017

  7. [7]

    Filmy cloud removal on satellite imagery with multispectral conditional generative adversarial nets,

    K. Enomoto, K. Sakurada, W. Wang, H. Fukui, M. Matsuoka, R. Naka- mura, and N. Kawaguchi, “Filmy cloud removal on satellite imagery with multispectral conditional generative adversarial nets,” in Proceed- ings of the IEEE conference on computer vision and pattern recognition workshops, 2017, pp. 48–56. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, V...

    work page 2017

  8. [8]

    Synthe- sis of multispectral optical images from sar/optical multitemporal data using conditional generative adversarial networks,

    J. D. Bermudez, P. N. Happ, R. Q. Feitosa, and D. A. Oliveira, “Synthe- sis of multispectral optical images from sar/optical multitemporal data using conditional generative adversarial networks,” IEEE Geoscience and Remote Sensing Letters , vol. 16, no. 8, pp. 1220–1224, 2019

    work page 2019

Show all 58 references
  1. [9]

    Cloud removal in satellite images using spatiotemporal generative networks. arxiv 2019,

    V . Sarukkai, A. Jain, B. Uzkent, and S. Ermon, “Cloud removal in satellite images using spatiotemporal generative networks. arxiv 2019,” arXiv preprint arXiv:1912.06838

    2019 arXiv

  2. [10]

    Uncrtaints: Uncertainty quantification for cloud removal in optical satellite time series,

    P. Ebel, V . S. F. Garnot, M. Schmitt, J. D. Wegner, and X. X. Zhu, “Uncrtaints: Uncertainty quantification for cloud removal in optical satellite time series,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 2085–2095

    2023

  3. [11]

    Cloud removal for remote sensing imagery via spatial attention generative adversarial network,

    H. Pan, “Cloud removal for remote sensing imagery via spatial attention generative adversarial network,” arXiv preprint arXiv:2009.13015, 2020

    2009 arXiv

  4. [12]

    Generative image inpainting with contextual attention,

    J. Yu, Z. Lin, J. Yang, X. Shen, X. Lu, and T. S. Huang, “Generative image inpainting with contextual attention,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2018

    2018

  5. [13]

    A remote sensing image dataset for cloud removal,

    D. Lin, G. Xu, X. Wang, Y . Wang, X. Sun, and K. Fu, “A remote sensing image dataset for cloud removal,” arXiv preprint arXiv:1901.00600 , 2019

    1901 arXiv

  6. [14]

    Kriging interpolation method and its application in retrieval of modis aerosol optical depth,

    C. Yu, L. Chen, L. Su, M. Fan, and S. Li, “Kriging interpolation method and its application in retrieval of modis aerosol optical depth,” in 2011 19th international conference on geoinformatics . IEEE, 2011, pp. 1–6

    2011

  7. [15]

    Remote sensing image reconstruction using tensor ring completion and total variation,

    W. He, N. Yokoya, L. Yuan, and Q. Zhao, “Remote sensing image reconstruction using tensor ring completion and total variation,” IEEE Transactions on Geoscience and Remote Sensing , vol. 57, no. 11, pp. 8998–9009, 2019

    2019

  8. [16]

    Image completion approaches using the statistics of similar patches,

    K. He and J. Sun, “Image completion approaches using the statistics of similar patches,” IEEE transactions on pattern analysis and machine intelligence, vol. 36, no. 12, pp. 2423–2435, 2014

    2014

  9. [17]

    A new method for retrieving band 6 of aqua modis,

    L. Wang, J. J. Qu, X. Xiong, X. Hao, Y . Xie, and N. Che, “A new method for retrieving band 6 of aqua modis,” IEEE Geoscience and Remote Sensing Letters , vol. 3, no. 2, pp. 267–270, 2006

    2006

  10. [18]

    Restoration of aqua modis band 6 using histogram matching and local least squares fitting,

    P. Rakwatin, W. Takeuchi, and Y . Yasuoka, “Restoration of aqua modis band 6 using histogram matching and local least squares fitting,” IEEE Transactions on Geoscience and Remote Sensing , vol. 47, no. 2, pp. 613–627, 2008

    2008

  11. [19]

    Quantitative restoration for modis band 6 on aqua,

    I. Gladkova, M. D. Grossberg, F. Shahriar, G. Bonev, and P. Romanov, “Quantitative restoration for modis band 6 on aqua,” IEEE transactions on geoscience and remote sensing , vol. 50, no. 6, pp. 2409–2416, 2011

    2011

  12. [20]

    Thin cloud removal from optical remote sensing images using the noise-adjusted principal com- ponents transform,

    M. Xu, X. Jia, M. Pickering, and S. Jia, “Thin cloud removal from optical remote sensing images using the noise-adjusted principal com- ponents transform,” ISPRS Journal of Photogrammetry and Remote Sensing, vol. 149, pp. 215–225, 2019

    2019

  13. [21]

    Block- gaussian-mixture priors for hyperspectral denoising and inpainting,

    A. M. Teodoro, J. M. Bioucas-Dias, and M. A. Figueiredo, “Block- gaussian-mixture priors for hyperspectral denoising and inpainting,” IEEE Transactions on Geoscience and Remote Sensing , vol. 59, no. 3, pp. 2478–2486, 2020

    2020

  14. [22]

    Reconstructing modis lst based on multitemporal classification and robust regression,

    C. Zeng, H. Shen, M. Zhong, L. Zhang, and P. Wu, “Reconstructing modis lst based on multitemporal classification and robust regression,” IEEE Geoscience and Remote Sensing Letters , vol. 12, no. 3, pp. 512– 516, 2014

    2014

  15. [23]

    A changing- weight filter method for reconstructing a high-quality ndvi time series to preserve the integrity of vegetation phenology,

    W. Zhu, Y . Pan, H. He, L. Wang, M. Mou, and J. Liu, “A changing- weight filter method for reconstructing a high-quality ndvi time series to preserve the integrity of vegetation phenology,” IEEE Transactions on Geoscience and Remote Sensing , vol. 50, no. 4, pp. 1085–1094, 2011

    2011

  16. [24]

    Removing clouds and recovering ground observations in satellite image sequences via temporally contiguous robust matrix completion,

    J. Wang, P. A. Olsen, A. R. Conn, and A. C. Lozano, “Removing clouds and recovering ground observations in satellite image sequences via temporally contiguous robust matrix completion,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2016, pp...

    2016

  17. [25]

    Recovering quantitative remote sensing products contaminated by thick clouds and shadows using multitemporal dictionary learning,

    X. Li, H. Shen, L. Zhang, H. Zhang, Q. Yuan, and G. Yang, “Recovering quantitative remote sensing products contaminated by thick clouds and shadows using multitemporal dictionary learning,” IEEE Transactions on Geoscience and Remote Sensing , vol. 52, no. 11, pp. 7086–7098, 2014

    2014

  18. [26]

    Thick cloud removal in landsat images based on autoregression of landsat time-series data,

    R. Cao, Y . Chen, J. Chen, X. Zhu, and M. Shen, “Thick cloud removal in landsat images based on autoregression of landsat time-series data,” Remote Sensing of Environment , vol. 249, p. 112001, 2020

    2020

  19. [27]

    Cloud removal for remotely sensed images by similar pixel replacement guided with a spatio-temporal mrf model,

    Q. Cheng, H. Shen, L. Zhang, Q. Yuan, and C. Zeng, “Cloud removal for remotely sensed images by similar pixel replacement guided with a spatio-temporal mrf model,” ISPRS journal of photogrammetry and remote sensing, vol. 92, pp. 54–68, 2014

    2014

  20. [28]

    Contextual spatiospectral postrecon- struction of cloud-contaminated images,

    S. Benabdelkader and F. Melgani, “Contextual spatiospectral postrecon- struction of cloud-contaminated images,” IEEE Geoscience and remote sensing letters, vol. 5, no. 2, pp. 204–208, 2008

    2008

  21. [29]

    Sparse-based reconstruction of missing information in remote sensing images from spectral/temporal complementary information,

    X. Li, H. Shen, L. Zhang, and H. Li, “Sparse-based reconstruction of missing information in remote sensing images from spectral/temporal complementary information,” ISPRS journal of photogrammetry and remote sensing, vol. 106, pp. 1–15, 2015

    2015

  22. [30]

    Spatially and temporally weighted regression: A novel method to produce continuous cloud- free landsat imagery,

    B. Chen, B. Huang, L. Chen, and B. Xu, “Spatially and temporally weighted regression: A novel method to produce continuous cloud- free landsat imagery,” IEEE Transactions on Geoscience and Remote Sensing, vol. 55, no. 1, pp. 27–37, 2016

    2016

  23. [31]

    Spatiotemporal reconstruction of land surface temperature derived from fengyun geosta- tionary satellite data,

    Z. Liu, P. Wu, S. Duan, W. Zhan, X. Ma, and Y . Wu, “Spatiotemporal reconstruction of land surface temperature derived from fengyun geosta- tionary satellite data,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 10, no. 10, pp. 4531–4543, 2017

    2017

  24. [32]

    Nonlocal tensor completion for multitemporal remotely sensed images’ inpainting,

    T.-Y . Ji, N. Yokoya, X. X. Zhu, and T.-Z. Huang, “Nonlocal tensor completion for multitemporal remotely sensed images’ inpainting,”IEEE Transactions on Geoscience and Remote Sensing , vol. 56, no. 6, pp. 3047–3061, 2018

    2018

  25. [33]

    Reconstructing cloud- contaminated multispectral images with contextualized autoencoder neu- ral networks,

    S. Malek, F. Melgani, Y . Bazi, and N. Alajlan, “Reconstructing cloud- contaminated multispectral images with contextualized autoencoder neu- ral networks,” IEEE Transactions on Geoscience and Remote Sensing , vol. 56, no. 4, pp. 2270–2282, 2017

    2017

  26. [34]

    Missing data reconstruction in remote sensing image with a unified spatial–temporal– spectral deep convolutional neural network,

    Q. Zhang, Q. Yuan, C. Zeng, X. Li, and Y . Wei, “Missing data reconstruction in remote sensing image with a unified spatial–temporal– spectral deep convolutional neural network,” IEEE Transactions on Geoscience and Remote Sensing , vol. 56, no. 8, pp. 4274–4288, 2018

    2018

  27. [35]

    Reconstructing geostation- ary satellite land surface temperature imagery based on a multiscale fea- ture connected convolutional neural network,

    P. Wu, Z. Yin, H. Yang, Y . Wu, and X. Ma, “Reconstructing geostation- ary satellite land surface temperature imagery based on a multiscale fea- ture connected convolutional neural network,” Remote Sensing, vol. 11, no. 3, p. 300, 2019

    2019

  28. [36]

    Thin cloud removal with residual symmetrical concatenation network,

    W. Li, Y . Li, D. Chen, and J. C.-W. Chan, “Thin cloud removal with residual symmetrical concatenation network,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 153, pp. 137–150, 2019

    2019

  29. [37]

    Thick clouds removal from multitemporal zy-3 satellite images using deep learning,

    Y . Chen, L. Tang, X. Yang, R. Fan, M. Bilal, and Q. Li, “Thick clouds removal from multitemporal zy-3 satellite images using deep learning,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 13, pp. 143–153, 2019

    2019

  30. [38]

    Thick cloud and cloud shadow removal in multitemporal imagery using progres- sively spatio-temporal patch group deep learning,

    Q. Zhang, Q. Yuan, J. Li, Z. Li, H. Shen, and L. Zhang, “Thick cloud and cloud shadow removal in multitemporal imagery using progres- sively spatio-temporal patch group deep learning,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 162, pp. 148–160, 2020

    2020

  31. [39]

    Deep internal learning for inpainting of cloud-affected regions in satellite imagery,

    M. Czerkawski, P. Upadhyay, C. Davison, A. Werkmeister, J. Cardona, R. Atkinson, C. Michie, I. Andonovic, M. Macdonald, and C. Tachtatzis, “Deep internal learning for inpainting of cloud-affected regions in satellite imagery,” Remote Sensing, vol. 14, no. 6, p. 1342, 2022

    2022

  32. [40]

    High-resolution cloud removal with multi-modal and multi-resolution data fusion: A new baseline and benchmark,

    F. Xu, Y . Shi, P. Ebel, W. Yang, and X. X. Zhu, “High-resolution cloud removal with multi-modal and multi-resolution data fusion: A new baseline and benchmark,” arXiv preprint arXiv:2301.03432 , 2023

    2023 arXiv

  33. [41]

    Cloud-gan: Cloud removal for sentinel-2 imagery using a cyclic consistent generative adversarial networks,

    P. Singh and N. Komodakis, “Cloud-gan: Cloud removal for sentinel-2 imagery using a cyclic consistent generative adversarial networks,” in IGARSS 2018-2018 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2018, pp. 1772–1775

    2018

  34. [42]

    A conditional generative adversarial network to fuse sar and multispectral optical data for cloud removal from sentinel-2 images,

    C. Grohnfeldt, M. Schmitt, and X. Zhu, “A conditional generative adversarial network to fuse sar and multispectral optical data for cloud removal from sentinel-2 images,” in IGARSS 2018-2018 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2018, pp. 1726–1729

    2018

  35. [43]

    Restoration of sea surface temperature satellite images using a partially occluded training set,

    S. Shibata, M. Iiyama, A. Hashimoto, and M. Minoh, “Restoration of sea surface temperature satellite images using a partially occluded training set,” in 2018 24th International Conference on Pattern Recognition (ICPR). IEEE, 2018, pp. 2771–2776

    2018

  36. [44]

    Sar to optical image synthesis for cloud removal with generative adversarial networks,

    J. Bermudez, P. Happ, D. Oliveira, and R. Feitosa, “Sar to optical image synthesis for cloud removal with generative adversarial networks,” ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 4, pp. 5–11, 2018

    2018

  37. [45]

    Cloud removal with fusion of high resolution optical and sar images using generative adversarial networks,

    J. Gao, Q. Yuan, J. Li, H. Zhang, and X. Su, “Cloud removal with fusion of high resolution optical and sar images using generative adversarial networks,” Remote Sensing, vol. 12, no. 1, p. 191, 2020

    2020

  38. [46]

    Glf-cr: Sar-enhanced cloud removal with global–local fusion,

    F. Xu, Y . Shi, P. Ebel, L. Yu, G.-S. Xia, W. Yang, and X. X. Zhu, “Glf-cr: Sar-enhanced cloud removal with global–local fusion,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 192, pp. 268–278, 2022

    2022

  39. [47]

    Ctgan: Cloud transformer generative adversarial network,

    G.-L. Huang and P.-Y . Wu, “Ctgan: Cloud transformer generative adversarial network,” in 2022 IEEE International Conference on Image Processing (ICIP). IEEE, 2022, pp. 511–515

    2022

  40. [48]

    Pmaa: A progres- sive multi-scale attention autoencoder model for high-performance cloud removal from multi-temporal satellite imagery,

    X. Zou, K. Li, J. Xing, P. Tao, and Y . Cui, “Pmaa: A progres- sive multi-scale attention autoencoder model for high-performance cloud removal from multi-temporal satellite imagery,” arXiv preprint arXiv:2303.16565, 2023

    2023 arXiv

  41. [49]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems , vol. 30, pp. 5998–6008, 2017. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 14, NO....

    2017

  42. [50]

    Rectified linear units improve restricted boltz- mann machines,

    V . Nair and G. E. Hinton, “Rectified linear units improve restricted boltz- mann machines,” in Proceedings of the 27th international conference on machine learning (ICML-10) , 2010, pp. 807–814

    2010

  43. [51]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2016, pp. 770–778

    2016

  44. [52]

    Root mean square error (rmse) or mean absolute error (mae)?–arguments against avoiding rmse in the literature,

    T. Chai and R. R. Draxler, “Root mean square error (rmse) or mean absolute error (mae)?–arguments against avoiding rmse in the literature,” Geoscientific model development , vol. 7, no. 3, pp. 1247–1250, 2014

    2014

  45. [53]

    Peak signal-to-noise ratio revisited: Is sim- ple beautiful?

    J. Korhonen and J. You, “Peak signal-to-noise ratio revisited: Is sim- ple beautiful?” in 2012 Fourth International Workshop on Quality of Multimedia Experience. IEEE, 2012, pp. 37–38

    2012

  46. [54]

    Image quality assessment: from error visibility to structural similarity,

    Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, “Image quality assessment: from error visibility to structural similarity,” IEEE transactions on image processing , vol. 13, no. 4, pp. 600–612, 2004

    2004

  47. [55]

    The spectral image processing system (sips)—interactive visualization and analysis of imaging spectrometer data,

    F. A. Kruse, A. Lefkoff, J. Boardman, K. Heidebrecht, A. Shapiro, P. Barloon, and A. Goetz, “The spectral image processing system (sips)—interactive visualization and analysis of imaging spectrometer data,” Remote sensing of environment , vol. 44, no. 2-3, pp. 145–163, 1993

    1993

  48. [56]

    Context adaptive network for image inpainting,

    Y . Deng, S. Hui, S. Zhou, W. Huang, and J. Wang, “Context adaptive network for image inpainting,” IEEE Transactions on Image Processing, 2023

    2023

  49. [57]

    Restormer: Efficient transformer for high-resolution image restoration,

    S. W. Zamir, A. Arora, S. Khan, M. Hayat, F. S. Khan, and M.-H. Yang, “Restormer: Efficient transformer for high-resolution image restoration,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 5728–5739. Wenli Huang received her P...

    2022

  50. [2018]

    degree in the Institute of Artificial Intelligence and Robotics at Xi’an Jiaotong University

    She is currently pursuing a Ph.D. degree in the Institute of Artificial Intelligence and Robotics at Xi’an Jiaotong University. Her research interests include Knowledge graph completion and Graph Representation Learning. Jinjun Wang received the B.E. and M.E. degrees from the ...

    2000

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.