Pith. sign in

REVIEW 5 major objections 5 minor 69 references

DACA-Net: A Degradation-Aware Conditional Diffusion Network for Underwater Image Enhancement

T0 review · 5 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read DACA-Net claims that conditioning a diffusion denoiser on a continuous, PSNR-derived degradation score yields state-of-the-art underwater image enhancement: 28.60 dB PSNR on UIEB and 28.57 dB on LSUI, with ablations showing that the…

desk verdict The degradation score that drives the whole pipeline needs the reference image at inference, so the method as written cannot be applied to a real underwater photo. read the letter →

arxiv 2507.22501 v1 pith:QNJT2AW6 submitted 2025-07-30 cs.CV eess.IV

Chang Huang , Jiahang Cao , Jun Ma , Kieren Yu , Cong Li , Huayong Yang , Kaishun Wu This is my paper · ORCID
classification cs.CVeess.IV
keywords underwaterimageenhancementconditionaldiffusionmodeldegradationscoreadaptivegroupnormalizationphysical-guidedfusionPSNRregressionSwinUNethybridloss
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 tries to establish that underwater image enhancement can be made adaptive by feeding the diffusion denoiser a continuous score that says how badly the input image is degraded. It claims that this score, regressed from the PSNR between a degraded image and its reference, lets the same network handle diverse water conditions by rescaling the noise schedule and modulating feature normalization through a module called AdaGN. If true, the approach would improve on prior state of the art on UIEB and LSUI, raising PSNR to 28.60 dB and 28.57 dB respectively while also improving no-reference quality scores. A sympathetic reading is that degradation-aware conditioning, not just a bigger diffusion backbone, is what carries the improvement.

What carries the argument

The load-bearing object is the degradation score D, a scalar in [0,1] computed as 1 minus the normalized PSNR between a degraded underwater image and its clean reference. D is injected into a conditional DDPM in two places: it rescales the noise schedule via beta_t * (1 + alpha D), and it drives a two-layer MLP that produces scale and shift parameters for Adaptive Group Normalisation (AdaGN), a variant of GroupNorm in which the per-group mean/variance normalization is modulated by the conditioning signal. The denoiser is a Swin-UNet backbone, and a Physical-guided Fusion Module (PGFM) amplifies the red channel by a degradation-dependent factor and applies frequency-domain attention. Together these components make the denoising trajectory depend on how degraded the input is, which is the mechanism the paper credits for adaptive enhancement.

What would settle it

Run the trained DACA-Net on a held-out underwater image with the reference withheld and trace whether the degradation estimator can produce D without the reference; if Eq. (1) is used as written, the conditioning signal is unavailable at inference and the claimed adaptive mechanism cannot operate. Alternatively, ablate D to a constant during denoising: if PSNR does not drop measurably, the degradation score is not carrying the improvement.

Watch

Extended reading notes

Core claim

The central claim is that DACA-Net, by conditioning a DDPM on an estimated degradation score D, achieves state-of-the-art underwater image enhancement. The score D is a normalized PSNR-derived scalar in [0,1] regressed from paired raw and reference images by a dual-stream CNN; it enters the denoising network twice, as a multiplier on the noise variance beta_t and as the input to an Adaptive Group Normalisation (AdaGN) layer that shifts and scales feature statistics. A Physical-guided Fusion Module adds red-channel amplification and frequency-domain attention. On UIEB the method reports PSNR 28.60 dB / SSIM 0.9456 and on LSUI 28.57 dB / 0.9344, ahead of the compared methods including GUPDM and DM_underwater; ablations attribute the gain to the joint action of AdaGN, PGFM, and the hybrid histogram-perceptual-contrastive loss.

Load-bearing premise

The degradation score that steers the whole enhancement is defined as a function of both the degraded image and its clean reference, and the paper does not describe how to compute it from a single raw image at test time.

Editorial extensions

If this is right

  • On the paper's evidence, replacing a fixed diffusion noise schedule with one scaled by a predicted degradation score improves PSNR and SSIM on UIEB and LSUI over fixed-schedule diffusion baselines.
  • The same conditioning scheme carries over to new datasets without retraining the score estimator if the PSNR range for normalization is known or estimated.
  • The degradation estimator is cheap enough (1.75 ms, 0.27 GFLOPs) that the adaptive conditioning adds little latency, so the framework is deployable near real time on a single GPU.
  • Because D is a continuous PSNR-based label, the network's behavior can be probed by construction: feeding a higher D should force stronger denoising and more aggressive red-channel compensation.

Reading between the lines

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

  • The reference-dependent definition of D (Eq. 1) means the training-time and test-time inputs of the estimator differ; a natural extension the paper does not spell out is a fully no-reference degradation regressor, which would make the method applicable to unpaired real footage.
  • The adaptive noise schedule beta_t * (1 + alpha D) could be tested in isolation: if alpha is learned to be near zero on some datasets, the claimed adaptivity is not actually engaged on those data.
  • The same degradation-conditioning recipe is transferable to other restoration domains with a scalar severity label, such as haze, rain, or low light, and the paper's ablations suggest the score, not the backbone, is the main driver.
  • Comparing DACA-Net against itself with D set to a constant would directly quantify how much of the 28.60 dB comes from degradation awareness rather than from the Swin-UNet and hybrid loss.
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

5 major / 5 minor

Summary. DACA-Net proposes a degradation-aware conditional diffusion model for underwater image enhancement. A dual-stream estimator regresses a scalar degradation score D from the degraded image and its reference, supervised by normalized PSNR; D is injected into a Swin-UNet diffusion backbone through an adaptive noise schedule, AdaGN, and a physical-guided fusion module. Training uses a hybrid loss of perceptual, histogram, and contrastive terms. On UIEB and LSUI the paper reports improvements over physics-based and learning-based baselines. The central problem is that D is defined on paired inputs, so the method is not deployable at inference as described, and the diffusion training objective is not specified.

Significance. If the proposed conditioning mechanism worked, a continuous degradation score could provide a principled way to adapt diffusion-based restoration to varying underwater degradations. The paper does not ship code or trained models, and all claimed results come from a single run without error bars, so the empirical contribution is not independently verifiable. The degradation-score idea is plausible, but the manuscript does not establish a working instance of it.

major comments (5)
  1. [Section 3.2, Eq. (1)] The degradation score is defined as D = R(phi(I_raw), phi(I_ref)), and Section 3.1 states it is regressed 'from paired raw and reference images.' No test-time procedure is given for computing D from the raw image alone, even though the abstract and Figure 1 describe the raw image as the input. Since D modulates the noise schedule (Eq. 5), AdaGN (Eq. 6), and PGFM red-channel gain (Eq. 7), the method as written cannot be deployed on real underwater images; the headline results in Tables 3-5 therefore cannot be reproduced from the paper alone.
  2. [Section 3.3, Eqs. (3)-(4)] Standard DDPM training assumes x_T ~ N(0,I) and a forward noising process on clean images. The paper instead states that 'the input is a raw underwater image, which is embedded as x_T and denoised,' but it never specifies the forward noising scheme, the noise-prediction objective, or any training loss for the reverse process from a non-Gaussian start state. Without this objective, the claimed conditional diffusion mechanism is undefined.
  3. [Section 3.5, Eq. (12)] The total loss L_total consists only of perceptual, histogram, and contrastive terms; there is no diffusion denoising loss (e.g., an epsilon-prediction or variational bound term). It is therefore unclear what objective trains the denoising network epsilon_theta, and the paper does not explain how the 'progressive denoising' contributes beyond the three loss terms.
  4. [Section 4.3.3, Table 3] The text states that DACA-Net obtains the highest UCIQE on UIEB, but the table lists DM_underwater with UCIQE 0.7230 versus DACA-Net's 0.7138, so the claim is contradicted by the paper's own numbers. More generally, all reported numbers are from a single run with no error bars, seeds, or code, which is insufficient support for the state-of-the-art claim.
  5. [Section 4.4.1, Table 5] The ablation narrative reports that removing all three loss components yields PSNR 15.08, but no row in Table 5 contains that value, and the row formatting makes the component configuration unreadable. The claimed baseline and the ablation conclusions are thus not verifiable.
minor comments (5)
  1. [Table 1] The parameter count column is labeled 'Param.(G)' and reports 1.02 for the proposed estimator; if the unit is gigaparameters this contradicts the 'lightweight' claim, and if it is megaparameters the header should be corrected.
  2. [Eq. (8)] The symbol F is used both for the feature map and for the FFT operator, making the formula difficult to parse.
  3. [Section 3.2, Eq. (2)] The direction of D should be stated explicitly: because D = 1 - normalized PSNR, higher D corresponds to stronger degradation; this is not obvious when it appears as a multiplicative factor in Eqs. (5) and (7).
  4. [Table 3 caption] The caption says the method achieves 'best or second-best performance,' but the text claims it is best on all metrics; on UIEB UCIQE the table shows DM_underwater ahead of DACA-Net.
  5. [Section 4.3.1] The inference-time measurement of the estimator is not accompanied by the hardware or resolution at which the time is measured; please specify the experimental conditions.

Circularity Check

1 steps flagged · score 6.0 of 10

The degradation score driving DACA-Net's conditioning is defined from the reference image (Eqs. 1-2), so the claimed 'prediction' is a fitted PSNR-based label rather than a raw-image-only estimate; reported gains rely on reference-derived conditioning, and no test-time substitute is given.

  1. fitted input called prediction [Section 3.1-3.2, Eq. (1)-(2); used in Eq. (5)-(7) and Table 3]
    "we first design a lightweight dual-stream convolutional network to regress a degradation score D from paired raw and reference images, supervised by the PSNR. ... D = R(phi(I_raw), phi(I_ref)) (1) ... D_label = 1 - (PSNR - PSNR_min)/(PSNR_max - PSNR_min) (2)"

    The conditioning signal D is defined as a function of both I_raw and I_ref, and its regression labels are a normalized PSNR computed between the degraded image and the reference. The paper presents D as a 'predicted degradation level' from the raw image, but by construction no raw-image-only estimate is defined or evaluated; a test-time substitute is never described. These D values enter the adaptive noise schedule, AdaGN, and PGFM, so the central mechanism is conditioned on the target reference. On the paired UIEB/LSUI test sets the reference is available, so the reported PSNR/SSIM/UIQM results can use reference-derived D, which is unavailable in real deployment.

full rationale

The paper's central claim is that degradation-aware conditional diffusion with the score D yields state-of-the-art underwater enhancement. The score is not, by the paper's own equations, a prediction from the raw image alone: Eq. (1) consumes both I_raw and I_ref, and Eq. (2) labels D by the PSNR of that same pair. The abstract and Section 3.3 state that D is predicted from the input raw image, yet no inference-time surrogate or retraining scheme is described for computing D from I_raw alone. Thus, on the paired benchmarks of Table 3, the conditioning signal can be obtained by leaking the reference into the estimator, which invalidates the claim that the gains come from a raw-only 'degradation awareness'. The diffusion network, Swin UNet backbone, and hybrid losses remain independent engineering components, and the final image is not literally equal to D, so the circularity is partial rather than total. There is no load-bearing self-citation or imported uniqueness theorem; the issue is that the key conditioning variable is fitted to a target-derived label and then called a prediction. This warrants a score of 6 rather than a lower score because the main claimed innovation is not instantiable as stated and the benchmark advantage can be attributed to reference-derived conditioning.

Assumptions & free parameters 5 free parameters · 6 assumptions · 1 invented entities

The central claim rests on an undefined inference-time degradation score, a diffusion start state that is not pure noise, and dataset-specific PSNR normalization; these are fitted or assumed rather than derived.

free parameters (5)
  • alpha = learned, value not reported
    Scales the degradation score in the adaptive noise schedule (Eq. 5).
  • gamma = learned, value not reported
    Scales the red-channel compensation in PGFM (Eq. 7).
  • loss weights lambda_1, lambda_2, lambda_3 = not reported
    Weights of the hybrid loss (Eq. 12), chosen by hand.
  • PSNR_min and PSNR_max per dataset = dataset-specific, values not reported
    Normalize PSNR to [0,1] for degradation labels (Eq. 2); fitted to each dataset.
  • number of diffusion timesteps T = 1500
    Set to 1500 in the implementation; an arbitrary hyperparameter.
assumptions (6)
  • standard math DDPM forward and reverse processes (Eqs. 3-4) are valid for the conditional restoration setting.
    Borrowed from Ho et al.; assumed without modification despite the nonstandard start state.
  • domain assumption Underwater degradation is caused primarily by wavelength-dependent absorption and scattering with strong red-channel attenuation.
    Used to justify the red-channel amplification in PGFM (Introduction, Eq. 7).
  • domain assumption PSNR is a valid scalar proxy for perceptual degradation and can serve as ground truth for D.
    Eq. 2 labels D with normalized PSNR; no perceptual validation is given.
  • ad hoc to paper The raw underwater image can be placed directly at x_T and denoised to a clean image.
    Section 3.3 states the raw image is embedded as x_T, but no forward process or training objective is defined for this mapping.
  • ad hoc to paper The degradation score D is available at inference even though Eq. (1) requires I_ref.
    No test-time substitute for the reference branch is described.
  • domain assumption UIEB references, obtained by human selection among algorithmic outputs, are reliable ground truth.
    Dataset description in Section 4.1; the ground truth is subjective.
invented entities (1)
  • Degradation score D
    purpose: Global conditioning signal controlling noise schedule, AdaGN feature modulation, and red-channel gain.
    A scalar regressed from paired raw and reference images to normalized PSNR; no falsifiable handle outside training, and its inference-time computation is unspecified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DACA-Net: A Degradation-Aware Conditional Diffusion Network for Underwater Image Enhancement." pith.science (2026). https://pith.science/paper/QNJT2AW6

@misc{pith2026250722501,
  author       = {Pith},
  title        = {Pith review of: DACA-Net: A Degradation-Aware Conditional Diffusion Network for Underwater Image Enhancement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QNJT2AW6}},
  note         = {Machine review of arXiv:2507.22501}
}
read the original abstract

Underwater images typically suffer from severe colour distortions, low visibility, and reduced structural clarity due to complex optical effects such as scattering and absorption, which greatly degrade their visual quality and limit the performance of downstream visual perception tasks. Existing enhancement methods often struggle to adaptively handle diverse degradation conditions and fail to leverage underwater-specific physical priors effectively. In this paper, we propose a degradation-aware conditional diffusion model to enhance underwater images adaptively and robustly. Given a degraded underwater image as input, we first predict its degradation level using a lightweight dual-stream convolutional network, generating a continuous degradation score as semantic guidance. Based on this score, we introduce a novel conditional diffusion-based restoration network with a Swin UNet backbone, enabling adaptive noise scheduling and hierarchical feature refinement. To incorporate underwater-specific physical priors, we further propose a degradation-guided adaptive feature fusion module and a hybrid loss function that combines perceptual consistency, histogram matching, and feature-level contrast. Comprehensive experiments on benchmark datasets demonstrate that our method effectively restores underwater images with superior colour fidelity, perceptual quality, and structural details. Compared with SOTA approaches, our framework achieves significant improvements in both quantitative metrics and qualitative visual assessments.

Figures

Figures reproduced from arXiv: 2507.22501 by the authors.

Figure 1
Figure 1. Overview illustration of the proposed DACA-Net framework. DACA-Net is composed of four main components: Dual [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Visual comparison with different methods on real-world datasets (UIEB and LSUI). The underwater images and [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Visualisation for ablation study of loss function. [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

69 extracted references · 53 canonical work pages

  1. [1]

    Ahmad Shahrizan Abdul Ghani and Nor Ashidi Mat Isa. 2014. Underwater image quality enhancement through composition of dual-intensity images and Rayleigh-stretching. SpringerPlus 3 (2014), 1–14

    work page 2014

  2. [2]

    Mahmoud Afifi and Michael S Brown. 2019. What else can fool deep learning? Addressing color constancy errors on deep neural network performance. In Proceedings of the IEEE/CVF international conference on computer vision . 243–252

    work page 2019

  3. [3]

    Derya Akkaynak and Tali Treibitz. 2018. A revised underwater image forma- tion model. In Proceedings of the IEEE conference on computer vision and pattern recognition. 6723–6732

    work page 2018

  4. [4]

    Derya Akkaynak and Tali Treibitz. 2019. Sea-thru: A method for removing water from underwater images. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 1682–1691

    work page 2019

  5. [5]

    Andrew Brock, Jeff Donahue, and Karen Simonyan. 2018. Large scale GAN training for high fidelity natural image synthesis. arXiv preprint arXiv:1809.11096 (2018)

    arXiv 2018

  6. [6]

    Hu Cao, Yueyue Wang, Joy Chen, Dongsheng Jiang, Xiaopeng Zhang, Qi Tian, and Manning Wang. 2022. Swin-unet: Unet-like pure transformer for medical image segmentation. In European conference on computer vision . Springer, 205–218

    work page 2022

  7. [7]

    Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. 2020. A simple framework for contrastive learning of visual representations. In Interna- tional conference on machine learning . PmLR, 1597–1607

    2020

  8. [8]

    John Y Chiang and Ying-Ching Chen. 2011. Underwater image enhancement by wavelength compensation and dehazing. IEEE transactions on image processing 21, 4 (2011), 1756–1769

    work page 2011

Show all 69 references
  1. [9]

    François Chollet. 2017. Xception: Deep learning with depthwise separable con- volutions. In Proceedings of the IEEE conference on computer vision and pattern recognition. 1251–1258

    2017

  2. [10]

    Roberto Danovaro, Cinzia Corinaldesi, Antonio Dell’Anno, and Paul VR Snelgrove

  3. [11]

    Evan N Edinger and Michael J Risk. 2000. Reef classification by coral morphology predicts coral reef conservation value. Biological Conservation 92, 1 (2000), 1–13

    2000

  4. [12]

    Cameron Fabbri, Md Jahidul Islam, and Junaed Sattar. 2018. Enhancing under- water imagery using generative adversarial networks. In 2018 IEEE international conference on robotics and automation (ICRA) . IEEE, 7159–7165

    2018

  5. [13]

    Renata Ferrari, David McKinnon, Hu He, Ryan N Smith, Peter Corke, Manuel González-Rivero, Peter J Mumby, and Ben Upcroft. 2016. Quantifying multiscale habitat structural complexity: a cost-effective framework for underwater 3D modelling. Remote Sensing 8, 2 (2016), 113

    2016

  6. [14]

    Zhenqi Fu, Huangxing Lin, Yan Yang, Shu Chai, Liyan Sun, Yue Huang, and Xing- hao Ding. 2022. Unsupervised underwater image restoration: From a homology perspective. In Proceedings of the AAAI conference on artificial intelligence , Vol. 36. 643–651

    2022

  7. [15]

    Leon A Gatys, Alexander S Ecker, and Matthias Bethge. 2015. A neural algorithm of artistic style. arXiv preprint arXiv:1508.06576 (2015)

    2015 arXiv

  8. [16]

    Ian J Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2014. Generative adversarial nets. Advances in neural information processing systems 27 (2014)

    2014

  9. [17]

    Meisheng Guan, Haiyong Xu, Gangyi Jiang, Mei Yu, Yeyao Chen, Ting Luo, and Xuebo Zhang. 2023. DiffWater: Underwater image enhancement based on conditional denoising diffusion probabilistic model. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing ...

    2023

  10. [18]

    Kaiming He, Jian Sun, and Xiaoou Tang. 2010. Single image haze removal using dark channel prior. IEEE transactions on pattern analysis and machine intelligence 33, 12 (2010), 2341–2353

    2010

  11. [19]

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition . 770–778

    2016

  12. [20]

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. Advances in neural information processing systems 33 (2020), 6840–6851

    2020

  13. [21]

    Ove Hoegh-Guldberg, Ken Caldeira, Thierry Chopin, Steve Gaines, Peter Hau- gan, Mark Hemer, Jennifer Howard, Manaswita Konar, Dorte Krause-Jensen, Catherine E Lovelock, et al. 2023. The ocean as a solution to climate change: five opportunities for action. In The blue compendiu...

    2023

  14. [22]

    Andrew G Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, and Hartwig Adam. 2017. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861 (2017)

    2017 arXiv

  15. [23]

    Dongmei Huang, Yan Wang, Wei Song, Jean Sequeira, and Sébastien Mavroma- tis. 2018. Shallow-water image enhancement using relative global histogram stretching based on adaptive parameter acquisition. In MultiMedia Modeling: 24th International Conference, MMM 2018, Bangkok, Tha...

    2018

  16. [24]

    Terry P Hughes, James T Kerry, Mariana Álvarez-Noriega, Jorge G Álvarez- Romero, Kristen D Anderson, Andrew H Baird, Russell C Babcock, Maria Beger, David R Bellwood, Ray Berkelmans, et al. 2017. Global warming and recurrent mass bleaching of corals. Nature 543, 7645 (2017), 373–377

    2017

  17. [25]

    Nils Gunnar Jerlov. 1976. Marine optics. Vol. 14. Elsevier

    1976

  18. [26]

    Daniel J Jobson, Zia-ur Rahman, and Glenn A Woodell. 1997. A multiscale retinex for bridging the gap between color images and the human observation of scenes. IEEE Transactions on Image processing 6, 7 (1997), 965–976

    1997

  19. [27]

    Justin Johnson, Alexandre Alahi, and Li Fei-Fei. 2016. Perceptual losses for real- time style transfer and super-resolution. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceed- ings, Part II 14 . Springer, 694–711

    2016

  20. [28]

    John TO Kirk. 1994. Light and photosynthesis in aquatic ecosystems . Cambridge university press

    1994

  21. [29]

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. 2012. Imagenet classifi- cation with deep convolutional neural networks. Advances in neural information processing systems 25 (2012)

    2012

  22. [30]

    Vladimir Kulikov, Shahar Yadin, Matan Kleiner, and Tomer Michaeli. 2023. Sinddm: A single image denoising diffusion model. In International conference on machine learning. PMLR, 17920–17930

    2023

  23. [31]

    Anat Levin, Dani Lischinski, and Yair Weiss. 2007. A closed-form solution to natural image matting. IEEE transactions on pattern analysis and machine intelligence 30, 2 (2007), 228–242

    2007

  24. [32]

    Chongyi Li, Saeed Anwar, Junhui Hou, Runmin Cong, Chunle Guo, and Wenqi Ren. 2021. Underwater image enhancement via medium transmission-guided multi-color space embedding. IEEE Transactions on Image Processing 30 (2021), 4985–5000

    2021

  25. [33]

    Chongyi Li, Saeed Anwar, and Fatih Porikli. 2020. Underwater scene prior inspired deep underwater image and video enhancement. Pattern recognition 98 (2020), 107038

    2020

  26. [34]

    Chongyi Li, Chunle Guo, Wenqi Ren, Runmin Cong, Junhui Hou, Sam Kwong, and Dacheng Tao. 2019. An underwater image enhancement benchmark dataset and beyond. IEEE transactions on image processing 29 (2019), 4376–4389

    2019

  27. [35]

    Chongyi Li and Jichang Guo. 2015. Underwater image enhancement by dehazing and color correction. Journal of Electronic Imaging 24, 3 (2015), 033023–033023

    2015

  28. [36]

    Jie Li, Katherine A Skinner, Ryan M Eustice, and Matthew Johnson-Roberson

  29. [37]

    Tongkun Liu, Bing Li, Xiao Du, Bingke Jiang, Leqi Geng, Feiyang Wang, and Zhuo Zhao. 2023. Fair: Frequency-aware image restoration for industrial visual anomaly detection. arXiv preprint arXiv:2309.07068 (2023)

    2023 arXiv

  30. [38]

    IEEE Robotics and Automation letters 3, 1 (2017), 387–394

    WaterGAN: Unsupervised generative network to enable real-time color correction of monocular underwater images. IEEE Robotics and Automation letters 3, 1 (2017), 387–394

    2017

  31. [39]

    Andreas Lugmayr, Martin Danelljan, Andres Romero, Fisher Yu, Radu Timofte, and Luc Van Gool. 2022. Repaint: Inpainting using denoising diffusion proba- bilistic models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 11461–11471

    2022

  32. [40]

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. 2021. Swin transformer: Hierarchical vision transformer us- ing shifted windows. In Proceedings of the IEEE/CVF international conference on computer vision. 10012–10022

    2021

  33. [41]

    Brian B Moser, Arundhati S Shanbhag, Federico Raue, Stanislav Frolov, Sebastian Palacio, and Andreas Dengel. 2024. Diffusion models, image super-resolution, and everything: A survey. IEEE Transactions on Neural Networks and Learning Systems (2024)

    2024

  34. [42]

    Delphine Mallet and Dominique Pelletier. 2014. Underwater video techniques for observing coastal marine biodiversity: a review of sixty years of publications (1952–2012). Fisheries Research 154 (2014), 44–62

    2014

  35. [43]

    David O Obura, Greta Aeby, Natchanon Amornthammarong, Ward Appeltans, Nicholas Bax, Joe Bishop, Russell E Brainard, Samuel Chan, Pamela Fletcher, Tim- othy AC Gordon, et al. 2019. Coral reef monitoring, reef assessment technologies, and ecosystem-based management. Frontiers in...

    2019

  36. [44]

    Pan Mu, Hanning Xu, Zheyuan Liu, Zheng Wang, Sixian Chan, and Cong Bai

  37. [45]

    Lintao Peng, Chunli Zhu, and Liheng Bian. 2023. U-shape transformer for un- derwater image enhancement. IEEE Transactions on Image Processing 32 (2023), 3066–3079

    2023

  38. [46]

    Yan-Tsung Peng and Pamela C Cosman. 2017. Underwater image restoration based on image blurriness and light absorption. IEEE transactions on image processing 26, 4 (2017), 1579–1594

    2017

  39. [47]

    Karen Panetta, Chen Gao, and Sos Agaian. 2015. Human-visual-system-inspired underwater image quality measures. IEEE Journal of Oceanic Engineering 41, 3 (2015), 541–551

    2015

  40. [48]

    Smitha Raveendran, Mukesh D Patil, and Gajanan K Birajdar. 2021. Underwater image enhancement: a comprehensive review, recent trends, challenges and applications. Artificial Intelligence Review 54 (2021), 5413–5467

    2021

  41. [49]

    Erik Reinhard, Michael Adhikhmin, Bruce Gooch, and Peter Shirley. 2001. Color transfer between images. IEEE Computer graphics and applications 21, 5 (2001), 34–41

    2001

  42. [50]

    Robin M Pope and Edward S Fry. 1997. Absorption spectrum (380–700 nm) of pure water. II. Integrating cavity measurements. Applied optics 36, 33 (1997), 8710–8723

    1997

  43. [51]

    Karen Simonyan and Andrew Zisserman. 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556 (2014)

    2014 arXiv

  44. [52]

    Kihyuk Sohn. 2016. Improved deep metric learning with multi-class n-pair loss objective. Advances in neural information processing systems 29 (2016)

    2016

  45. [53]

    Subham Sahoo, Aaron Gokaslan, Christopher M De Sa, and Volodymyr Kuleshov

  46. [54]

    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. 2020. Score-based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456 (2020)

    2020 arXiv

  47. [55]

    Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott Reed, Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, and Andrew Rabinovich. 2015. Going deeper with convolutions. InProceedings of the IEEE conference on computer vision and pattern recognition . 1–9

    2015

  48. [56]

    Yi Tang, Hiroshi Kawasaki, and Takafumi Iwaguchi. 2023. Underwater image enhancement by transformer-based diffusion model with non-uniform sampling for skip strategy. In Proceedings of the 31st ACM international conference on multimedia. 5419–5427

    2023

  49. [57]

    Wei Song, Yan Wang, Dongmei Huang, and Dian Tjondronegoro. 2018. A rapid scene depth estimation model based on underwater light attenuation prior for underwater image restoration. In Advances in Multimedia Information Processing– PCM 2018: 19th Pacific-Rim Conference on Multim...

    2018

  50. [58]

    Ning Wang, Tingkai Chen, Xiangjun Kong, Yanzheng Chen, Rongfeng Wang, Yongjun Gong, and Shiji Song. 2023. Underwater attentional generative adver- sarial networks for image enhancement. IEEE Transactions on Human-Machine Systems 53, 3 (2023), 490–500

    2023

  51. [59]

    Zhizhong Wang, Lei Zhao, and Wei Xing. 2023. Stylediffusion: Controllable disentangled style transfer via diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 7677–7689

    2023

  52. [60]

    Yuxin Wu and Kaiming He. 2018. Group normalization. In Proceedings of the European conference on computer vision (ECCV) . 3–19

    2018

  53. [61]

    Derek P Tittensor, Camilo Mora, Walter Jetz, Heike K Lotze, Daniel Ricard, Edward Vanden Berghe, and Boris Worm. 2010. Global patterns and predictors of marine biodiversity across taxa. Nature 466, 7310 (2010), 1098–1101

    2010

  54. [62]

    Dehuan Zhang, Jingchun Zhou, Chunle Guo, Weishi Zhang, and Chongyi Li. 2024. Synergistic multiscale detail refinement via intrinsic supervision for underwater image enhancement. InProceedings of the AAAI conference on artificial intelligence, Vol. 38. 7033–7041

    2024

  55. [63]

    Tingting Zhang, Yujie Li, and Shinya Takahashi. 2021. Underwater image en- hancement using improved generative adversarial network. Concurrency and Computation: Practice and Experience 33, 22 (2021), e5841

    2021

  56. [64]

    Xiangyu Zhang, Xinyu Zhou, Mengxiao Lin, and Jian Sun. 2018. Shufflenet: An ex- tremely efficient convolutional neural network for mobile devices. In Proceedings of the IEEE conference on computer vision and pattern recognition . 6848–6856

    2018

  57. [65]

    Miao Yang and Arcot Sowmya. 2015. An underwater color image quality evalua- tion metric. IEEE Transactions on Image Processing 24, 12 (2015), 6062–6071

    2015

  58. [69]

    Hang Zhao, Orazio Gallo, Iuri Frosio, and Jan Kautz. 2016. Loss functions for image restoration with neural networks. IEEE Transactions on computational imaging 3, 1 (2016), 47–57

    2016

  59. [2017]

    Current Biology 27, 11 (2017), R461– R465

    The deep-sea under global change. Current Biology 27, 11 (2017), R461– R465

    2017

  60. [2023]

    In Proceedings of the 31st ACM international conference on multimedia

    A generalized physical-knowledge-guided dynamic model for underwater image enhancement. In Proceedings of the 31st ACM international conference on multimedia. 7111–7120

  61. [2024]

    Advances in Neural Informa- tion Processing Systems 37 (2024), 105730–105779

    Diffusion models with learned adaptive noise. Advances in Neural Informa- tion Processing Systems 37 (2024), 105730–105779

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.