Pith. sign in

REVIEW 3 major objections 4 minor 30 references

CMAMRNet: A Contextual Mask-Aware Network Enhancing Mural Restoration Through Comprehensive Mask Guidance

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

Pith's one-line read The paper claims CMAMRNet, a mask-aware U-shaped transformer with Mask-Aware Up/Down-Samplers and Co-Feature Aggregators, outperforms twelve baselines on Dunhuang mural restoration benchmarks.

desk verdict Solid mask-aware inpainting architecture with a real benchmark gap: the SOTA claim doesn't land until mural-specific baselines and variance are reported. read the letter →

arxiv 2508.07140 v2 pith:4QHCRHD2 submitted 2025-08-10 cs.CV

classification cs.CV
keywords muralrestorationimageinpaintingmask-awarenetworktransformerMAUDSCo-FeatureAggregatorDunhuangmuralsculturalheritage
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 argues that mural restoration improves when the deterioration mask is kept explicit at every resolution transition rather than being fed once at the input. It proposes CMAMRNet, a U-shaped transformer where Mask-Aware Up/Down-Samplers carry the mask through pixel-shuffle and channel-interleaving paths at every scale change, and Co-Feature Aggregators combine channel- and spatial-attention features at the finest and coarsest scales. On the MuralDH and Dunhuang benchmarks, the full model reports the best PSNR, SSIM, MAE, and LPIPS among twelve baselines, surpassing the previous best method. If these benchmark results hold, the practical payoff is a restoration pipeline that keeps attention pinned to genuinely damaged regions while preserving fine artistic structure, which matters for digital cultural-heritage conservation.

What carries the argument

The load-bearing machinery is the pair of modules the paper names Mask-Aware Up/Down-Sampler (MAUDS) and Co-Feature Aggregator (CFA). MAUDS sits at every resolution transition in the U-shaped transformer: the Down-Sampler pixel-unshuffles the binary mask into four channels and interleaves those mask channels between feature channels before a depthwise convolution, while the Up-Sampler pixel-shuffles the mask and adds it to the odd channels of the upsampled features. CFA runs in parallel at the highest and lowest resolutions, passing image-plus-mask through two MaxViT blocks and then through Channel and Spatial Feature Focusing Blocks that use FFT to capture frequency patterns; the mask branc

What would settle it

Run CMAMRNet on murals whose damage masks come from a different annotator or an automated crack detector while keeping training data fixed; if PSNR and SSIM drop sharply relative to the dataset's own masks, then the gains depend on the oracle mask distribution rather than on general mask guidance. A stronger test is expert evaluation of restorations against a known historical intact state, where synthetic-target metrics provide no ground truth.

Watch

Extended reading notes

Core claim

CMAMRNet's central claim is that persistent mask guidance, not just mask-conditioned input or loss weighting, is what makes mural inpainting work. The network maintains a binary deterioration mask alongside image features through every down- and up-sampling step, so the mask cannot be diluted by convolutions. The Mask-Aware Down-Sampler interleaves mask channels with selected feature channels after pixel unshuffle, while the Mask-Aware Up-Sampler adds a pixel-shuffled mask to odd-channel-selected features before depthwise refinement. At the coarsest and finest scales, the Co-Feature Aggregator uses parallel channel- and spatial-focusing blocks with FFT to modulate features by a mask-derived

Load-bearing premise

The load-bearing premise is that the deterioration masks and intact ground-truth images in the MuralDH and Dunhuang datasets faithfully represent real mural damage and authentic restoration targets; if the masks are synthetic or unrepresentative, the reported benchmark gains measure inpainting on artificial masks rather than restoration of real murals.

Editorial extensions

If this is right

  • On MuralDH, the complete model reaches PSNR 36.2565 and SSIM 0.9683, beating the previous best method by about 0.77 dB in PSNR while also lowering MAE and LPIPS.
  • On Dunhuang, it reaches PSNR 33.0738 and SSIM 0.8944, a larger relative gain over the prior best than on MuralDH, suggesting mask guidance is especially valuable on the harder benchmark.
  • Ablations show that removing MAUDS drops PSNR by about 0.81 dB and raises MAE and LPIPS, while removing CFA also degrades perceptual quality, indicating both components contribute and work complementarily.
  • Because the loss is only MSE plus SSIM and training uses 256x256 crops, the reported gains come from architecture rather than heavy supervision, implying the module design may transfer to other image-restoration settings.

Reading between the lines

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

  • An implication the authors leave implicit is that MAUDS propagates whatever mask it is given, so on real murals the quality of the damage annotation may matter as much as network capacity; a natural test is feeding masks from an automated crack detector and measuring the drop.
  • The channel-interleaving mechanism in the Mask-Aware Down-Sampler is generic: it could transfer to other image-to-image tasks with a binary prior, such as shadow removal, reflection suppression, or document cleanup.
  • The paper's quantitative metrics cannot fully capture historical authenticity, so a testable extension is expert art-historian evaluation of restored mural details against real conservation standards.
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 / 4 minor

Summary. The paper proposes CMAMRNet, a U-shaped Transformer network for digital mural restoration that combines two new components: MAUDS (Mask-Aware Up/Down-Sampler), which propagates mask information through resolution changes via channel selection and mask-guided fusion, and CFA (Co-Feature Aggregator), which fuses channel, spatial, and frequency features at the highest and lowest resolutions. The model is trained with an MSE+SSIM loss and evaluated on the MuralDH and Dunhuang mural datasets. The authors report consistent improvements over twelve general image-inpainting baselines on PSNR, SSIM, MAE, and LPIPS, with the largest gains over HINT, and they present ablations showing that both MAUDS and CFA contribute to the final performance. The code is publicly released.

Significance. If the reported results are robust, the central idea—maintaining explicit mask guidance through every sampling stage rather than letting masks attenuate through the network—is a sensible and potentially useful contribution to cultural-heritage restoration. The architecture is clearly described, the ablations indicate that both proposed modules help, and the release of code is a practical strength for reproducibility. However, the headline claim of state-of-the-art mural restoration is not yet substantiated: the comparison omits mural-specific methods cited in the paper, and no uncertainty quantification or statistical tests are provided. As it stands, the paper demonstrates a well-engineered inpainting method on two mural benchmarks, but the stronger claim of superiority over prior mural-restoration approaches requires additional evidence.

major comments (3)
  1. [§3.3, Table 1] The quantitative comparison includes only general image-inpainting methods. The Introduction explicitly discusses mural-specific restoration techniques, including line-drawing guided inpainting [7], Thanka mural inpainting [17], color restoration [22], and MuralDiff [24], the last of which is critiqued at length. None of these appears in Table 1. Therefore, the statement that HINT is the 'previous state-of-the-art' and that CMAMRNet surpasses it is not established for mural restoration. The reported margins over HINT are modest (0.77 dB PSNR on MuralDH, 0.83 dB on Dunhuang), so adding MuralDiff or other mural-specific baselines could change the ranking. A direct comparison is required before the central SOTA claim can be accepted.
  2. [§3.2, §3.3, Tables 1–2] All quantitative results appear to come from a single training run, with no error bars, multiple seeds, or significance tests. The differences over HINT are only about 0.8 dB PSNR, which may be within run-to-run variability for deep inpainting models. Similarly, the ablation differences in Table 2 (e.g., row (1) vs. row (3)) need confidence intervals. Please report mean±std over at least three seeds for the main comparisons and ablations, and preferably a paired significance test, so readers can judge whether the reported gains are reliable.
  3. [§3.1] The paper does not specify how the deterioration masks in MuralDH and Dunhuang are generated or annotated, nor whether they represent real mural damage patterns. If the masks are synthetic or based on simplified stroke simulation, the improvements may reflect performance on artificial degradation rather than authentic mural restoration. Please clarify the mask provenance in the datasets, and ideally include an evaluation on real damage masks or discuss the extent to which the benchmark masks are representative. This is important for the external validity of the central claim that CMAMRNet improves mural restoration specifically.
minor comments (4)
  1. [§2.2, Figure 2] The caption of Figure 2 labels (b) as 'Mask-Aware Down-Sampler (MADS)' and (c) as 'Mask-Aware Up-Sampler (MAUS)', but the text states that the MAUS upsampling is 'As illustrated in Fig. 2 (b)' and the MADS downsampling is 'As depicted in Fig. 2 (c)'. These references are reversed; please correct the figure or the text.
  2. [§2.2, Eq. (7)] The interleaving notation in Eq. (7) is difficult to follow: the use of 'Finter,2i−1' and 'Finter,2i' to denote individual channels, with the mask index 'j = i mod 4', is unconventional. Please define the channel indexing more explicitly, e.g., with explicit loops or a clearer channel-assembly description.
  3. [General] Minor typographical and style issues: 'Corresponging author' in the footnote, 'GeLu' in Figure 2 should be 'GELU', 'Misf' in Table 1 should be 'MISF', and the figure label 'EdgeConn' should be expanded consistently with the main text.
  4. [§3.2] The use of 'mix-up' in data augmentation is mentioned without detail. It is unclear how mix-up is applied when a corresponding mask is required. Please specify whether the mask is also mixed, or clarify the augmentation protocol.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claim is an empirical benchmark comparison, and all load-bearing components are independently defined and tested on held-out data.

full rationale

The paper's central claim is empirical: CMAMRNet outperforms prior methods on the MuralDH and Dunhuang benchmarks (Table 1). This is supported by held-out test-set metrics, not derived from the model's own definitions. The two novel components, MAUDS and CFA, are specified by explicit equations (Eqs. 3-12) operating on input features Fimg and mask M; their outputs are not defined in terms of the reported PSNR/SSIM results. The loss function combines MSE and SSIM with λ = 0.4, which is an empirically set hyperparameter; tuning a loss weight on the training/validation split is standard practice and does not constitute fitting a prediction to its own target. The ablation study (Table 2) further isolates the contribution of each module. The paper cites several works by its own authors ([1], [2], [21], and [24] MuralDiff), but these citations are used as related-work examples or as motivation for the proposed approach, not as load-bearing evidence for the headline result. The absence of MuralDiff from Table 1 is a potential completeness concern about the SOTA claim, but it is a missing-baseline issue, not a circularity. No self-definitional step, no fitted-parameter-renamed-as-prediction, no imported uniqueness theorem, and no ansatz-smuggled-via-citation pattern is present. The derivation is therefore self-contained relative to its empirical validation.

Assumptions & free parameters 4 free parameters · 4 assumptions · 2 invented entities

The central performance claim depends on tuned hyperparameters such as loss weight and training resolution, on dataset assumptions about mask fidelity and baseline fairness, and on the new modules MAUDS and CFA whose validity is established only by in-paper ablations.

free parameters (4)
  • Loss weight lambda for SSIM = 0.4
    Chosen empirically in Section 2.4. It balances pixel-level and structural losses and directly affects the reported metrics.
  • Training patch size = 256x256
    Hand-chosen. Testing at 512x512 while training at 256x256 creates a resolution gap that can affect comparisons.
  • Initial learning rate = 2e-4
    Standard AdamW schedule. Needed to reproduce training but not fitted to the data.
  • Batch size = 4
    Set by hardware. Reproducibility depends on this value.
assumptions (4)
  • domain assumption The MuralDH and Dunhuang masks are accurate pixel-level annotations of deterioration and their intact images are authentic restoration targets (Section 3.1).
    The entire evaluation rests on these datasets representing real mural damage. No independent validation of mask fidelity is provided.
  • domain assumption The 12 compared baselines were trained and evaluated under the same protocol as CMAMRNet (Sections 3.2-3.3).
    The paper does not state the training details for baselines, so the PSNR gaps assume a fair comparison.
  • domain assumption PSNR, SSIM, MAE, and LPIPS are accepted proxies for restoration quality and artistic authenticity (Section 3.1).
    These are standard metrics but do not measure historical or artistic correctness.
  • standard math Restormer blocks, MaxViT, and attention mechanisms function as described in the cited works (Section 2).
    The architecture reuses published modules without formal proof.
invented entities (2)
  • MAUDS (Mask-Aware Up/Down-Sampler)
    purpose: Maintain mask information across resolution changes in the encoder and decoder.
    The module's benefit is shown only in the paper's ablation on one dataset, with no external verification.
  • CFA (Co-Feature Aggregator) with CFFB and SFFB
    purpose: Combine fine texture and global structure features at the highest and lowest resolutions.
    Only internal ablations on MuralDH support its effectiveness; no independent benchmark.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CMAMRNet: A Contextual Mask-Aware Network Enhancing Mural Restoration Through Comprehensive Mask Guidance." pith.science (2026). https://pith.science/paper/4QHCRHD2

@misc{pith2026250807140,
  author       = {Pith},
  title        = {Pith review of: CMAMRNet: A Contextual Mask-Aware Network Enhancing Mural Restoration Through Comprehensive Mask Guidance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4QHCRHD2}},
  note         = {Machine review of arXiv:2508.07140}
}
read the original abstract

Murals, as invaluable cultural artifacts, face continuous deterioration from environmental factors and human activities. Digital restoration of murals faces unique challenges due to their complex degradation patterns and the critical need to preserve artistic authenticity. Existing learning-based methods struggle with maintaining consistent mask guidance throughout their networks, leading to insufficient focus on damaged regions and compromised restoration quality. We propose CMAMRNet, a Contextual Mask-Aware Mural Restoration Network that addresses these limitations through comprehensive mask guidance and multi-scale feature extraction. Our framework introduces two key components: (1) the Mask-Aware Up/Down-Sampler (MAUDS), which ensures consistent mask sensitivity across resolution scales through dedicated channel-wise feature selection and mask-guided feature fusion; and (2) the Co-Feature Aggregator (CFA), operating at both the highest and lowest resolutions to extract complementary features for capturing fine textures and global structures in degraded regions. Experimental results on benchmark datasets demonstrate that CMAMRNet outperforms state-of-the-art methods, effectively preserving both structural integrity and artistic details in restored murals. The code is available at~\href{https://github.com/CXH-Research/CMAMRNet}{https://github.com/CXH-Research/CMAMRNet}.

Figures

Figures reproduced from arXiv: 2508.07140 by the authors.

Figure 1
Figure 1. Architecture overview of the proposed CMAMRNet. The framework is built upon [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Overview of Mask-Aware Up/Down-Sampler (MAUDS): (a) Sampling Block, (b) [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Architecture of: (a) the Channel Feature Focusing Block (CFFB), (b) the Spatial [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison with state-of-the-art methods. From left to right: (a) [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 29 canonical work pages

  1. [7]

    Line Drawing Guided Progressive Inpainting of Mural Damage

    Luxi Li, Qin Zou, Fan Zhang, Hongkai Yu, Long Chen, Chengfang Song, Xianfeng Huang, and Xiaoguang Wang. Line drawing guided progressive inpainting of mural damages. arXiv preprint arXiv:2211.06649, 2022

  2. [17]

    Thanka mural inpainting based on multi-scale adaptive partial convolution and stroke-like mask.IEEE Transactions on Image Processing, 2021

    Nianyi Wang, Weilan Wang, Wenjin Hu, Aaron Fenster, and Shuo Li. Thanka mural inpainting based on multi-scale adaptive partial convolution and stroke-like mask.IEEE Transactions on Image Processing, 2021

  3. [22]

    Deep learning algorithm in ancient relics image colour restoration technology

    Wanni Xu and Youlei Fu. Deep learning algorithm in ancient relics image colour restoration technology. Multimedia Tools and Applications, 2023

  4. [24]

    Muraldiff: Diffusion for ancient murals restoration on large-scale pre-training

    Zishan Xu, Xiaofeng Zhang, Wei Chen, Jueting Liu, Tingting Xu, and Zehua Wang. Muraldiff: Diffusion for ancient murals restoration on large-scale pre-training. IEEE Transactions on Emerging Topics in Computational Intelligence, 2024

  5. [1]

    Retinex- mamba: Retinex-based mamba for low-light image enhancement

    Jiesong Bai, Yuhao Yin, Qiyuan He, Yuanxian Li, and Xiaofeng Zhang. Retinex- mamba: Retinex-based mamba for low-light image enhancement. In International Conference on Neural Information Processing, pages 427–442. Springer, 2024

  6. [2]

    LensNet: An End-to-End Learning Framework for Empirical Point Spread Function Modeling and Lensless Imaging Reconstruction

    Jiesong Bai, Yuhao Yin, Yihang Dong, Xiaofeng Zhang, Chi-Man Pun, and Xuhang Chen. Lensnet: An end-to-end learning framework for empirical point spread function modeling and lensless imaging reconstruction.arXiv preprint arXiv:2505.01755, 2025

  7. [3]

    A scientific investigation of five polymeric materials used in the conservation of murals in dunhuang mogao grottoes

    Su Bomin, Zhang Huabing, Zhang Binjian, Jiang Deqiang, Zhang Rui, and Tan Xiang. A scientific investigation of five polymeric materials used in the conservation of murals in dunhuang mogao grottoes. Journal of Cultural Heritage, 2018

  8. [4]

    Hint: High-quality inpainting transformer with mask-aware encoding and enhanced attention.IEEE Trans- actions on Multimedia, 2024

    Shuang Chen, Amir Atapour-Abarghouei, and Hubert PH Shum. Hint: High-quality inpainting transformer with mask-aware encoding and enhanced attention.IEEE Trans- actions on Multimedia, 2024

Show all 30 references
  1. [5]

    T-former: An effi- cient transformer for image inpainting

    Ye Deng, Siqi Hui, Sanping Zhou, Deyu Meng, and Jinjun Wang. T-former: An effi- cient transformer for image inpainting. In Proceedings of the 30th ACM International Conference on Multimedia (ACM MM), 2022

  2. [6]

    Deep fusion network for image completion

    Xin Hong, Pengfei Xiong, Renhe Ji, and Haoqiang Fan. Deep fusion network for image completion. In Proceedings of the 27th ACM International Conference on Multimedia (ACM MM), 2019. LEI, ET AL: CMAMRNET 11

  3. [8]

    Misf: Multi- level interactive siamese filtering for high-fidelity image inpainting

    Xiaoguang Li, Qing Guo, Di Lin, Ping Li, Wei Feng, and Song Wang. Misf: Multi- level interactive siamese filtering for high-fidelity image inpainting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022

  4. [9]

    Coordfill: Efficient high-resolution image inpainting via parameterized coordi- nate querying

    Weihuang Liu, Xiaodong Cun, Chi-Man Pun, Menghan Xia, Yong Zhang, and Jue Wang. Coordfill: Efficient high-resolution image inpainting via parameterized coordi- nate querying. In Proceedings of the AAAI Conference on Artificial Intelligence, 2023

  5. [10]

    SGDR: Stochastic gradient descent with warm restarts

    Ilya Loshchilov and Frank Hutter. SGDR: Stochastic gradient descent with warm restarts. In International Conference on Learning Representations (ICLR), 2017

  6. [11]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. In Interna- tional Conference on Learning Representations (ICLR), 2019

  7. [12]

    The digital reconstruction of degraded ancient temple murals using dynamic mask generation and an extended exemplar-based region- filling algorithm

    V Rakhi Mol and P Uma Maheswari. The digital reconstruction of degraded ancient temple murals using dynamic mask generation and an extended exemplar-based region- filling algorithm. Heritage Science, 2021

  8. [13]

    Edge- connect: Structure guided image inpainting using edge prediction

    Kamyar Nazeri, Eric Ng, Tony Joseph, Faisal Qureshi, and Mehran Ebrahimi. Edge- connect: Structure guided image inpainting using edge prediction. In 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW), 2019

  9. [14]

    Distance-based weighted transformer network for image completion

    Pourya Shamsolmoali, Masoumeh Zareapoor, Huiyu Zhou, Xuelong Li, and Yue Lu. Distance-based weighted transformer network for image completion. Pattern Recogni- tion, 2024

  10. [15]

    Resolution-robust large mask inpainting with fourier convolutions

    Roman Suvorov, Elizaveta Logacheva, Anton Mashikhin, Anastasia Remizova, Arsenii Ashukha, Aleksei Silvestrov, Naejin Kong, Harshith Goka, Kiwoong Park, and Vic- tor Lempitsky. Resolution-robust large mask inpainting with fourier convolutions. In Proceedings of the IEEE/CVF Win...

  11. [16]

    Maxvit: Multi-axis vision transformer

    Zhengzhong Tu, Hossein Talebi, Han Zhang, Feng Yang, Peyman Milanfar, Alan Bovik, and Yinxiao Li. Maxvit: Multi-axis vision transformer. In Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2022

  12. [18]

    Current progress on murals: distribution, conservation and utilization

    Yihui Wang and Xiaodong Wu. Current progress on murals: distribution, conservation and utilization. Heritage Science, 2023

  13. [19]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: from error visibility to structural similarity. IEEE Transactions on Image Processing, 2004. 12 LEI, ET AL: CMAMRNET

  14. [20]

    Syformer: Structure-guided synergism transformer for large-portion image inpainting

    Jie Wu, Yuchao Feng, Honghui Xu, Chuanmeng Zhu, and Jianwei Zheng. Syformer: Structure-guided synergism transformer for large-portion image inpainting. In Pro- ceedings of the AAAI Conference on Artificial Intelligence, 2024

  15. [21]

    Dlen: Dual branch of transformer for low- light image enhancement in dual domains

    Junyu Xia, Jiesong Bai, and Yihang Dong. Dlen: Dual branch of transformer for low- light image enhancement in dual domains. arXiv preprint arXiv:2501.12235, 2025

  16. [23]

    A comprehensive dataset for digital restoration of dunhuang murals

    Zishan Xu, Yuqing Yang, Qianzhen Fang, Wei Chen, Tingting Xu, Jueting Liu, and Zehua Wang. A comprehensive dataset for digital restoration of dunhuang murals. Scientific Data, 2024

  17. [25]

    Contextual resid- ual aggregation for ultra high-resolution image inpainting

    Zili Yi, Qiang Tang, Shekoofeh Azizi, Daesik Jang, and Zhan Xu. Contextual resid- ual aggregation for ultra high-resolution image inpainting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020

  18. [26]

    Free- form image inpainting with gated convolution

    Jiahui Yu, Zhe Lin, Jimei Yang, Xiaohui Shen, Xin Lu, and Thomas S Huang. Free- form image inpainting with gated convolution. In Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision (ICCV), 2019

  19. [27]

    Region normalization for image inpainting

    Tao Yu, Zongyu Guo, Xin Jin, Shilin Wu, Zhibo Chen, Weiping Li, Zhizheng Zhang, and Sen Liu. Region normalization for image inpainting. In Proceedings of the AAAI Conference on Artificial Intelligence, 2020

  20. [28]

    Dunhuang grottoes painting dataset and benchmark.arXiv preprint arXiv:1907.04589, 2019

    Tianxiu Yu, Shijie Zhang, Cong Lin, Shaodi You, Jian Wu, Jiawan Zhang, Xiaohong Ding, and Huili An. Dunhuang grottoes painting dataset and benchmark.arXiv preprint arXiv:1907.04589, 2019

  21. [29]

    Restormer: Efficient transformer for high-resolution image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan, and Ming–Hsuan Yang. Restormer: Efficient transformer for high-resolution image restoration. In Proceedings of the Conference on Computer Vision and Pattern Recognition (CVPR), 2022

  22. [30]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2018

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.