pith. sign in

arxiv: 2605.17506 · v1 · pith:UAJOQEJLnew · submitted 2026-05-17 · 💻 cs.CV

Degradation Frequency Curve: An Explicit Frequency-Quantified Representation for All-in-One Image Restoration

Xinghua Huang , Zhixiong Yang , Chen Wu , Shengxi Li , Shuaifeng Zhi , Yue Zhang , Qibin Hou , Xin Deng
show 1 more author
Jingyuan Xia
This is my paper

Pith reviewed 2026-05-20 13:44 UTC · model grok-4.3

classification 💻 cs.CV
keywords image restorationdegradation representationfrequency domain analysisall-in-one restorationspectral tokensblind image restorationmixed degradation
0
0 comments X

The pith

The Degradation Frequency Curve turns image degradations into explicit, measurable spectral representations that guide all-in-one blind restoration.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

In all-in-one blind image restoration, degradation is usually handled only as an implicit part of the mapping from degraded to clean images. The paper instead treats degradation as an explicit object that can be measured directly. It does this by defining the Degradation Frequency Curve through band-wise residual-to-degraded energy ratios computed in the frequency domain. These ratios convert visually entangled effects into a coordinate space that decomposes into reusable spectral tokens. The tokens then condition a multi-scale restorer to handle mixed, compound, and unseen degradations more effectively.

Core claim

DFC quantifies degradation responses by measuring band-wise residual-to-degraded energy ratios in the frequency domain. This converts visually entangled and hard-to-describe degradation effects into a measurable degradation coordinate space. DFC can be adaptively decomposed into band-wise spectral tokens, allowing local degradation responses to be represented as reusable restoration priors. The resulting DFC-guided Image Restorer estimates these curves from intermediate results and uses the tokens to steer coarse-to-fine restoration.

What carries the argument

Degradation Frequency Curve (DFC), the structured spectral representation obtained from band-wise residual-to-degraded energy ratios in the frequency domain, which supplies both a measurable coordinate space and decomposable spectral tokens for restoration guidance.

If this is right

  • DFC supplies an explicit object that can be measured and manipulated instead of remaining hidden inside the restoration mapping.
  • Spectral tokens derived from DFC act as reusable priors that condition restoration at multiple scales.
  • The same representation supports progressive estimation of degradation from intermediate restorations.
  • Performance gains appear on standard, composite, unseen, and real-world benchmarks simultaneously.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The frequency-ratio approach could be tested on video sequences by adding a temporal-frequency axis to track motion-coupled degradations.
  • If the tokens prove stable across domains, they might serve as conditioning signals for generative models that synthesize degraded-clean pairs on demand.
  • Replacing hand-crafted priors in older restoration pipelines with learned DFC tokens offers a direct route to measure generalization improvements.

Load-bearing premise

Band-wise residual-to-degraded energy ratios in the frequency domain supply a sufficient and generalizable quantification of degradation effects even under mixed, compound, or unseen conditions.

What would settle it

No improvement or outright worse performance when DFC-IR is tested on a held-out set of real-world images whose degradations combine multiple unseen types in ways not captured by the energy-ratio measurements.

Figures

Figures reproduced from arXiv: 2605.17506 by Chen Wu, Jingyuan Xia, Qibin Hou, Shengxi Li, Shuaifeng Zhi, Xin Deng, Xinghua Huang, Yue Zhang, Zhixiong Yang.

Figure 1
Figure 1. Figure 1: Motivation of the proposed DFC-guided restoration paradigm. (a) Degraded images (from top to bottom: haze, rain, and noise); (b) The corresponding [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of DFC construction. The DFC is obtained by transforming the residual and degraded images into the frequency domain, computing [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Statistical DFC profiles for haze, rain, and noise. The solid lines and colored bands represent the mean profiles and variance bands, respectively. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Severity-aware DFC variation. (a)(b) DFC profiles and peak response variations under different Gaussian noise levels [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Illustration of DFC tokenization and compositional token priors for unseen degradations. (a) Seen degradation DFCs are decomposed into band-wise [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: The architecture of the proposed DFC-IR. (a) The overall multi-scale encoder-decoder framework. (b) Adaptive DFC tokenization implemented by [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Visual results for all-in-one restoration on five tasks. The pixel-wise error is visualized using a heatmap, where a color gradient from black to white [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visual comparisons on unseen restoration tasks, with UIE on UIEB shown in the first row and desnowing on CSD shown in the second row. Our [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Visual comparisons of generalization evaluation with five-degradation models on real-world NH-HAZE and SPANet. [PITH_FULL_IMAGE:figures/full_fig_p012_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Visual comparisons of generalization evaluation with five-degradation models on real-world composite degradations across haze/rain (row 1) and [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Visual results of uniform tokenization and our DFC-guided tok [PITH_FULL_IMAGE:figures/full_fig_p013_11.png] view at source ↗
read the original abstract

A fundamental difficulty in all-in-one blind image restoration is that degradation is usually treated as an implicit factor hidden in degraded-to-clean mapping, rather than as an explicit object that can be measured and manipulated. This limitation becomes more pronounced under mixed, compound, or unseen degradation conditions, where degradation effects are hard to assign to predefined labels or task-specific parameters. We propose the Degradation Frequency Curve (DFC), a structured spectral representation that quantifies degradation responses by measuring band-wise residual-to-degraded energy ratios in the frequency domain. DFC converts visually entangled and hard-to-describe degradation effects into a measurable degradation coordinate space. Moreover, DFC can be adaptively decomposed into band-wise spectral tokens, allowing local degradation responses to be represented as reusable restoration priors. Based on this representation, we develop the DFC-guided Image Restorer (DFC-IR), a token-conditioned multi-scale framework that progressively estimates DFCs from intermediate restorations and uses the resulting spectral tokens to guide degradation-aware restoration in a coarse-to-fine manner. Extensive experiments on standard, composite, unseen, and real-world degradation benchmarks show that DFC provides an effective representation basis for all-in-one restoration, leading to state-of-the-art performance and improved generalization under complex degradation profiles.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 3 minor

Summary. The paper introduces the Degradation Frequency Curve (DFC), defined via band-wise residual-to-degraded energy ratios in the frequency domain, as an explicit representation that converts entangled degradation effects into a measurable coordinate space for all-in-one blind image restoration. DFC is adaptively decomposed into band-wise spectral tokens serving as reusable priors. The authors present DFC-IR, a token-conditioned multi-scale network that progressively estimates DFCs from intermediate restorations to guide coarse-to-fine, degradation-aware restoration. Experiments report state-of-the-art results on standard, composite, unseen, and real-world benchmarks.

Significance. If the DFC representation and its progressive estimation prove robust, the work could advance all-in-one restoration by replacing implicit mappings with an explicit, frequency-quantified degradation space that supports generalization to mixed and unseen conditions. The adaptive token decomposition and multi-scale conditioning are technically interesting contributions, and the broad experimental coverage across degradation profiles provides a solid basis for evaluating the approach.

major comments (1)
  1. [§3] §3 (DFC-IR framework): the central claim that DFC supplies a sufficient and generalizable quantification for guiding restoration under compound or unseen degradations rests on progressive estimation of DFCs from intermediate restorations. Because residuals are defined relative to ground-truth clean images (available only in training), inference-time estimation risks error propagation; without a dedicated ablation or stability analysis (e.g., oracle DFC vs. estimated DFC on unseen mixed degradations), the generalization results are not yet fully supported.
minor comments (3)
  1. The abstract would be strengthened by including the explicit mathematical definition of the DFC (band-wise residual-to-degraded energy ratio) rather than describing it only in prose.
  2. [§2] Notation for frequency-band partitioning and the precise computation of energy ratios should be stated once in a dedicated equation or table for reproducibility.
  3. Figure captions and legends could more clearly indicate which curves correspond to which degradation types to aid visual interpretation of the DFC representation.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the thorough review and constructive feedback on our manuscript. We address the major comment below and outline the revisions we plan to make to strengthen the presentation of our results.

read point-by-point responses

  1. Referee: [§3] §3 (DFC-IR framework): the central claim that DFC supplies a sufficient and generalizable quantification for guiding restoration under compound or unseen degradations rests on progressive estimation of DFCs from intermediate restorations. Because residuals are defined relative to ground-truth clean images (available only in training), inference-time estimation risks error propagation; without a dedicated ablation or stability analysis (e.g., oracle DFC vs. estimated DFC on unseen mixed degradations), the generalization results are not yet fully supported.

    Authors: We agree that a dedicated analysis of the progressive DFC estimation at inference time would further support our claims. In the current manuscript, the DFC-IR is trained end-to-end, allowing the network to learn robust estimation from intermediate restorations. However, to directly address potential error propagation under compound and unseen degradations, we will add an ablation study comparing restoration performance when using oracle DFCs (computed using ground-truth clean images) versus the progressively estimated DFCs. This analysis will be conducted on the composite and unseen degradation benchmarks to quantify the stability of the estimation process and its impact on generalization. revision: yes

Circularity Check

0 steps flagged

DFC defined via direct residual energy ratios; no reduction to inputs by construction

full rationale

The paper defines the Degradation Frequency Curve explicitly as band-wise residual-to-degraded energy ratios measured in the frequency domain. This construction is a direct computation from available signals during training and does not derive the representation from fitted parameters, self-referential predictions, or prior self-citations. The progressive estimation of DFCs from intermediate restorations at inference is presented as an implementation detail for the DFC-IR framework rather than a load-bearing step that forces the core definition. No equations or claims in the provided text reduce the claimed representation to its own outputs by construction, and the central premise remains independent of the target restoration performance. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim depends on the domain assumption that frequency-domain energy ratios capture degradation in a way that can be decomposed into reusable tokens for restoration guidance. No free parameters or invented entities beyond the DFC itself are explicitly quantified in the abstract.

axioms (1)
  • domain assumption Degradation effects can be quantified by measuring band-wise residual-to-degraded energy ratios in the frequency domain.
    This forms the definition of DFC and is invoked as the basis for converting entangled degradations into a measurable space.
invented entities (1)
  • Degradation Frequency Curve (DFC) no independent evidence
    purpose: To serve as an explicit, measurable representation of degradation that can be decomposed into spectral tokens for guiding restoration.
    Newly proposed construct introduced to address limitations of implicit degradation handling; no independent evidence outside the paper is described in the abstract.

pith-pipeline@v0.9.0 · 5778 in / 1217 out tokens · 39522 ms · 2026-05-20T13:44:43.943940+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

59 extracted references · 59 canonical work pages · 2 internal anchors

  1. [1]

    Neighbor2neighbor: Self- supervised denoising from single noisy images,

    T. Huang, S. Li, X. Jia, H. Lu, and J. Liu, “Neighbor2neighbor: Self- supervised denoising from single noisy images,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 14 781–14 790

    work page 2021

  2. [2]

    Unsupervised image de- noising in real-world scenarios via self-collaboration parallel generative adversarial branches,

    X. Lin, C. Ren, X. Liu, J. Huang, and Y . Lei, “Unsupervised image de- noising in real-world scenarios via self-collaboration parallel generative adversarial branches,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 12 642–12 652

    work page 2023

  3. [3]

    Rethinking coarse-to-fine approach in single image deblurring,

    S.-J. Cho, S.-W. Ji, J.-P. Hong, S.-W. Jung, and S.-J. Ko, “Rethinking coarse-to-fine approach in single image deblurring,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 4641–4650

    work page 2021

  4. [4]

    Deblurgan-v2: De- blurring (orders-of-magnitude) faster and better,

    O. Kupyn, T. Martyniuk, J. Wu, and Z. Wang, “Deblurgan-v2: De- blurring (orders-of-magnitude) faster and better,” inProceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 8878– 8887

    work page 2019

  5. [5]

    Robust representation learning with feedback for single image deraining,

    C. Chen and H. Li, “Robust representation learning with feedback for single image deraining,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 7742–7751

    work page 2021

  6. [6]

    Multi-scale progressive fusion network for single image deraining,

    K. Jiang, Z. Wang, P. Yi, C. Chen, B. Huang, Y . Luo, J. Ma, and J. Jiang, “Multi-scale progressive fusion network for single image deraining,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 8346–8355

    work page 2020

  7. [7]

    All-in-one medical image restoration via task-adaptive routing,

    Z. Yang, H. Chen, Z. Qian, Y . Yi, H. Zhang, D. Zhao, B. Wei, and Y . Xu, “All-in-one medical image restoration via task-adaptive routing,” in International Conference on Medical Image Computing and Computer- Assisted Intervention. Springer, 2024, pp. 67–77

    work page 2024

  8. [8]

    Promptir: Prompting for all-in-one image restoration,

    V . Potlapalli, S. W. Zamir, S. H. Khan, and F. Shahbaz Khan, “Promptir: Prompting for all-in-one image restoration,”Advances in Neural Infor- mation Processing Systems, vol. 36, pp. 71 275–71 293, 2023

    work page 2023

  9. [9]

    Instructir: High-quality image restoration following human instructions,

    M. V . Conde, G. Geigle, and R. Timofte, “Instructir: High-quality image restoration following human instructions,” inEuropean Conference on Computer Vision. Springer, 2024, pp. 1–21

    work page 2024

  10. [10]

    All-in-one image restoration for unknown corruption,

    B. Li, X. Liu, P. Hu, Z. Wu, J. Lv, and X. Peng, “All-in-one image restoration for unknown corruption,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 17 452–17 462

    work page 2022

  11. [11]

    All-in-one image restoration for unknown degradations using adaptive discriminative filters for specific degradations,

    D. Park, B. H. Lee, and S. Y . Chun, “All-in-one image restoration for unknown degradations using adaptive discriminative filters for specific degradations,” in2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2023, pp. 5815–5824

    work page 2023

  12. [12]

    Smartassign: Learning a smart knowledge assignment strategy for deraining and desnowing,

    Y . Wang, C. Ma, and J. Liu, “Smartassign: Learning a smart knowledge assignment strategy for deraining and desnowing,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 3677–3686

    work page 2023

  13. [13]

    Gridformer: Residual dense transformer with grid structure for image restoration in adverse weather conditions,

    T. Wang, K. Zhang, Z. Shao, W. Luo, B. Stenger, T. Lu, T.-K. Kim, W. Liu, and H. Li, “Gridformer: Residual dense transformer with grid structure for image restoration in adverse weather conditions,” International journal of computer vision, vol. 132, no. 10, pp. 4541– 4563, 2024

    work page 2024

  14. [14]

    Complexity experts are task-discriminative learners for any image restoration,

    E. Zamfir, Z. Wu, N. Mehta, Y . Tan, D. P. Paudel, Y . Zhang, and R. Timofte, “Complexity experts are task-discriminative learners for any image restoration,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 12 753–12 763

    work page 2025

  15. [15]

    Adair: Adaptive all-in-one image restoration via frequency mining and modu- lation,

    Y . Cui, S. W. Zamir, S. Khan, A. Knoll, M. Shah, and F. S. Khan, “Adair: Adaptive all-in-one image restoration via frequency mining and modu- lation,” in13th International Conference on Learning Representations, ICLR 2025. International Conference on Learning Representations, ICLR, 2025, pp. 57 335–57 356

    work page 2025

  16. [16]

    Vision-language gradient descent-driven all-in-one deep unfolding networks,

    H. Zeng, X. Wang, Y . Chen, J. Su, and J. Liu, “Vision-language gradient descent-driven all-in-one deep unfolding networks,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 7524–7533. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 15

    work page 2025

  17. [17]

    Degradation-aware feature perturbation for all-in-one image restoration,

    X. Tian, X. Liao, X. Liu, M. Li, and C. Ren, “Degradation-aware feature perturbation for all-in-one image restoration,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 28 165– 28 175

    work page 2025

  18. [18]

    Onerestore: A universal restoration framework for composite degradation,

    Y . Guo, Y . Gao, Y . Lu, H. Zhu, R. W. Liu, and S. He, “Onerestore: A universal restoration framework for composite degradation,” inEuropean conference on computer vision. Springer, 2024, pp. 255–272

    work page 2024

  19. [19]

    Neural degradation representation learning for all-in-one image restoration,

    M. Yao, R. Xu, Y . Guan, J. Huang, and Z. Xiong, “Neural degradation representation learning for all-in-one image restoration,”IEEE transac- tions on image processing, vol. 33, pp. 5408–5423, 2024

    work page 2024

  20. [20]

    Promptrestorer: A prompting image restoration method with degrada- tion perception,

    C. Wang, J. Pan, W. Wang, J. Dong, M. Wang, Y . Ju, and J. Chen, “Promptrestorer: A prompting image restoration method with degrada- tion perception,”Advances in Neural Information Processing Systems, vol. 36, pp. 8898–8912, 2023

    work page 2023

  21. [21]

    arXiv preprint arXiv:2310.01018 , volume=

    Z. Luo, F. K. Gustafsson, Z. Zhao, J. Sj ¨olund, and T. B. Sch ¨on, “Controlling vision-language models for multi-task image restoration,” arXiv preprint arXiv:2310.01018, 2023

    work page arXiv 2023

  22. [22]

    Multimodal prompt perceiver: Empower adaptiveness generalizability and fidelity for all-in- one image restoration,

    Y . Ai, H. Huang, X. Zhou, J. Wang, and R. He, “Multimodal prompt perceiver: Empower adaptiveness generalizability and fidelity for all-in- one image restoration,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 25 432–25 444

    work page 2024

  23. [23]

    Uniprocessor: a text- induced unified low-level image processor,

    H. Duan, X. Min, S. Wu, W. Shen, and G. Zhai, “Uniprocessor: a text- induced unified low-level image processor,” inEuropean Conference on Computer Vision. Springer, 2024, pp. 180–199

    work page 2024

  24. [24]

    Uni- versal image restoration pre-training via degradation classi- fication.arXiv preprint arXiv:2501.15510, 2025

    J. Hu, L. Jin, Z. Yao, and Y . Lu, “Universal image restoration pre-training via degradation classification,”arXiv preprint arXiv:2501.15510, 2025

    work page arXiv 2025

  25. [25]

    Image restoration by estimating frequency distribution of local patches,

    J. Yoo, S.-h. Lee, and N. Kwak, “Image restoration by estimating frequency distribution of local patches,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 6684– 6692

    work page 2018

  26. [26]

    Learning in the frequency domain,

    K. Xu, M. Qin, F. Sun, Y . Wang, Y .-K. Chen, and F. Ren, “Learning in the frequency domain,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 1740–1749

    work page 2020

  27. [27]

    Focal frequency loss for image reconstruction and synthesis,

    L. Jiang, B. Dai, W. Wu, and C. C. Loy, “Focal frequency loss for image reconstruction and synthesis,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 13 919–13 929

    work page 2021

  28. [28]

    Selective frequency network for image restoration,

    Y . Cui, Y . Tao, Z. Bing, W. Ren, X. Gao, X. Cao, K. Huang, and A. Knoll, “Selective frequency network for image restoration,” inThe eleventh international conference on learning representations, 2023

    work page 2023

  29. [29]

    Image restoration via frequency selection,

    Y . Cui, W. Ren, X. Cao, and A. Knoll, “Image restoration via frequency selection,”IEEE Transactions on Pattern Analysis and Machine Intelli- gence, vol. 46, no. 2, pp. 1093–1108, 2023

    work page 2023

  30. [30]

    Perceive-ir: Learning to perceive degradation better for all-in-one image restoration,

    X. Zhang, J. Ma, G. Wang, Q. Zhang, H. Zhang, and L. Zhang, “Perceive-ir: Learning to perceive degradation better for all-in-one image restoration,”IEEE Transactions on Image Processing, 2025

    work page 2025

  31. [31]

    Ingredient-oriented multi-degradation learning for image restoration,

    J. Zhang, J. Huang, M. Yao, Z. Yang, H. Yu, M. Zhou, and F. Zhao, “Ingredient-oriented multi-degradation learning for image restoration,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 5825–5835

    work page 2023

  32. [32]

    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

    work page 2004

  33. [33]

    The unreasonable effectiveness of deep features as a perceptual metric,

    R. Zhang, P. Isola, A. A. Efros, E. Shechtman, and O. Wang, “The unreasonable effectiveness of deep features as a perceptual metric,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 586–595

    work page 2018

  34. [34]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium,

    M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, and S. Hochreiter, “Gans trained by a two time-scale update rule converge to a local nash equilibrium,”Advances in neural information processing systems, vol. 30, 2017

    work page 2017

  35. [35]

    Maniqa: Multi-dimension attention network for no-reference image quality assessment,

    S. Yang, T. Wu, S. Shi, S. Lao, Y . Gong, M. Cao, J. Wang, and Y . Yang, “Maniqa: Multi-dimension attention network for no-reference image quality assessment,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 1191–1200

    work page 2022

  36. [36]

    Exploring clip for assessing the look and feel of images,

    J. Wang, K. C. Chan, and C. C. Loy, “Exploring clip for assessing the look and feel of images,” inProceedings of the AAAI conference on artificial intelligence, vol. 37, no. 2, 2023, pp. 2555–2563

    work page 2023

  37. [37]

    Musiq: Multi- scale image quality transformer,

    J. Ke, Q. Wang, Y . Wang, P. Milanfar, and F. Yang, “Musiq: Multi- scale image quality transformer,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 5148–5157

    work page 2021

  38. [38]

    Contour detection and hierarchical image segmentation,

    P. Arbelaez, M. Maire, C. Fowlkes, and J. Malik, “Contour detection and hierarchical image segmentation,”IEEE transactions on pattern analysis and machine intelligence, vol. 33, no. 5, pp. 898–916, 2010

    work page 2010

  39. [39]

    Waterloo exploration database: New challenges for image quality assessment models,

    K. Ma, Z. Duanmu, Q. Wu, Z. Wang, H. Yong, H. Li, and L. Zhang, “Waterloo exploration database: New challenges for image quality assessment models,”IEEE Transactions on Image Processing, vol. 26, no. 2, pp. 1004–1016, 2016

    work page 2016

  40. [40]

    A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics,

    D. Martin, C. Fowlkes, D. Tal, and J. Malik, “A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics,” inProceedings eighth IEEE international conference on computer vision. ICCV 2001, vol. 2. IEEE, 2001, pp. 416–423

    work page 2001

  41. [41]

    Deep joint rain detection and removal from a single image,

    W. Yang, R. T. Tan, J. Feng, J. Liu, Z. Guo, and S. Yan, “Deep joint rain detection and removal from a single image,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 1357–1366

    work page 2017

  42. [42]

    Benchmarking single-image dehazing and beyond,

    B. Li, W. Ren, D. Fu, D. Tao, D. Feng, W. Zeng, and Z. Wang, “Benchmarking single-image dehazing and beyond,”IEEE transactions on image processing, vol. 28, no. 1, pp. 492–505, 2018

    work page 2018

  43. [43]

    Deep multi-scale convolutional neural network for dynamic scene deblurring,

    S. Nah, T. Hyun Kim, and K. Mu Lee, “Deep multi-scale convolutional neural network for dynamic scene deblurring,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 3883–3891

    work page 2017

  44. [44]

    Deep Retinex Decomposition for Low-Light Enhancement

    C. Wei, W. Wang, W. Yang, and J. Liu, “Deep retinex decomposition for low-light enhancement,”arXiv preprint arXiv:1808.04560, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  45. [45]

    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,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 5728–5739

    work page 2022

  46. [46]

    Adam: A Method for Stochastic Optimization

    D. P. Kingma, “Adam: A method for stochastic optimization,”arXiv preprint arXiv:1412.6980, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  47. [47]

    Deep generalized unfolding networks for image restoration,

    C. Mou, Q. Wang, and J. Zhang, “Deep generalized unfolding networks for image restoration,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 17 399–17 410

    work page 2022

  48. [48]

    Multi-stage progressive image restoration,

    S. W. Zamir, A. Arora, S. Khan, M. Hayat, F. S. Khan, M.-H. Yang, and L. Shao, “Multi-stage progressive image restoration,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 14 821–14 831

    work page 2021

  49. [49]

    Mambair: A simple baseline for image restoration with state-space model,

    H. Guo, J. Li, T. Dai, Z. Ouyang, X. Ren, and S.-T. Xia, “Mambair: A simple baseline for image restoration with state-space model,” in European conference on computer vision. Springer, 2024, pp. 222– 241

    work page 2024

  50. [50]

    Learning weather-general and weather-specific features for image restoration under multiple adverse weather conditions,

    Y . Zhu, T. Wang, X. Fu, X. Yang, X. Guo, J. Dai, Y . Qiao, and X. Hu, “Learning weather-general and weather-specific features for image restoration under multiple adverse weather conditions,” inPro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 21 747–21 758

    work page 2023

  51. [51]

    Restoring vision in adverse weather conditions with patch-based denoising diffusion models,

    O. ¨Ozdenizci and R. Legenstein, “Restoring vision in adverse weather conditions with patch-based denoising diffusion models,”IEEE trans- actions on pattern analysis and machine intelligence, vol. 45, no. 8, pp. 10 346–10 357, 2023

    work page 2023

  52. [52]

    Harmony in diversity: Improving all-in-one image restoration via multi-task collaboration,

    G. Wu, J. Jiang, K. Jiang, and X. Liu, “Harmony in diversity: Improving all-in-one image restoration via multi-task collaboration,” inProceedings of the 32nd ACM international conference on multimedia, 2024, pp. 6015–6023

    work page 2024

  53. [53]

    Degradation-aware residual-conditioned optimal transport for unified image restoration,

    X. Tang, X. Gu, X. He, X. Hu, and J. Sun, “Degradation-aware residual-conditioned optimal transport for unified image restoration,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    work page 2025

  54. [54]

    An underwater image enhancement benchmark dataset and beyond,

    C. Li, C. Guo, W. Ren, R. Cong, J. Hou, S. Kwong, and D. Tao, “An underwater image enhancement benchmark dataset and beyond,”IEEE transactions on image processing, vol. 29, pp. 4376–4389, 2019

    work page 2019

  55. [55]

    All snow removed: Single image desnowing algo- rithm using hierarchical dual-tree complex wavelet representation and contradict channel loss,

    W.-T. Chen, H.-Y . Fang, C.-L. Hsieh, C.-C. Tsai, I. Chen, J.-J. Ding, S.-Y . Kuoet al., “All snow removed: Single image desnowing algo- rithm using hierarchical dual-tree complex wavelet representation and contradict channel loss,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 4196–4205

    work page 2021

  56. [56]

    Nh-haze: An image dehazing benchmark with non-homogeneous hazy and haze-free images,

    C. O. Ancuti, C. Ancuti, and R. Timofte, “Nh-haze: An image dehazing benchmark with non-homogeneous hazy and haze-free images,” inPro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, 2020, pp. 444–445

    work page 2020

  57. [57]

    Spatial attentive single-image deraining with a high quality real rain dataset,

    T. Wang, X. Yang, K. Xu, S. Chen, Q. Zhang, and R. W. Lau, “Spatial attentive single-image deraining with a high quality real rain dataset,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 12 270–12 279

    work page 2019

  58. [58]

    Sparse gradient reg- ularized deep retinex network for robust low-light image enhancement,

    W. Yang, W. Wang, H. Huang, S. Wang, and J. Liu, “Sparse gradient reg- ularized deep retinex network for robust low-light image enhancement,” IEEE Transactions on Image Processing, vol. 30, pp. 2072–2086, 2021

    work page 2072

  59. [59]

    A com- parative study for single image blind deblurring,

    W.-S. Lai, J.-B. Huang, Z. Hu, N. Ahuja, and M.-H. Yang, “A com- parative study for single image blind deblurring,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 1701–1709

    work page 2016