Pith. sign in

REVIEW 4 major objections 5 minor 85 references

High-Quality Exposure Correction with Diffusion-Based Image Generation Priors

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Diffusion priors fix exposure correction in one denoising step, the paper argues.

desk verdict A plausible single-step diffusion-prior recipe for exposure correction, but the abstract's consistent-SOTA claim is contradicted by the paper's own tables and the missing artifacts keep the key gain unverified. read the letter →

arxiv 2608.08720 v1 pith:CZGQKS67 submitted 2026-08-09 cs.CV

Ziwen Li , Meng Cao , Jinpu Zhang , Chunyang Li , Long Bao , Heng Sun , Yuehuan Wang This is my paper · ORCID
classification cs.CV
keywords exposurecorrectiondiffusionmodelStablegenerativepriorsimageenhancementcross-attentionsingle-stepdenoising
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 claims that exposure correction can be recast as a single-step conditional denoising task by fine-tuning a pre-trained Stable Diffusion model, then fusing its low-frequency features into a regression network that preserves high-frequency detail. The method, DPEC, is reported to outperform prior state-of-the-art methods on the LCDP, MSEC, and SICE benchmarks in both fidelity and perceptual metrics. The key move is to let the diffusion model work only on a 64x64 downsampled image, so it supplies believable low-frequency content for blown-out or crushed regions without the artifacts and cost of full-resolution iterative sampling. A joint cross-attention module merges the diffusion and regression features at multiple scales. If the claims hold, DPEC offers a practical recipe for exploiting generative priors in image restoration at regression-like inference cost.

What carries the argument

The central object is the fine-tuned denoising U-Net used as a low-frequency prior generator. Because the U-Net input is the concatenation of the noisy latent and a 64x64 downsampled version of the input image, the model only needs to refine illumination and large-scale structure rather than synthesize textures; training is restricted to the maximum timestep so the network learns one noise level, and inference is a single DDIM-style step using the v-prediction parameterization. The Joint Cross-Attention Module (JCAM) then lets diffusion features and regression features each query a concatenated key-value set, giving symmetric fusion across scales.

What would settle it

Run DPEC on LCDP with the diffusion branch replaced by a trainable bilinear or VAE-based low-frequency generator at the same 64x64 resolution and identical fusion; if PSNR and LPIPS remain within noise of the full model, the generative prior itself is not contributing. Alternatively, feed the diffusion branch a zero image with no content and check whether the one-step output still resembles plausible image structure; if it does, the prior dominates the condition, which would predict failure on inputs whose extreme exposure regions cover most of the frame.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that a pre-trained diffusion model can be adapted to exposure correction by fine-tuning it to denoise at a single fixed timestep ($T=1000$) with the downsampled poor-exposure image concatenated as condition, enabling one-step generation; and that the resulting low-frequency prior, when fused into a regression network through a joint cross-attention module, yields enhanced images that are simultaneously more faithful and more perceptually natural than those from regression-only or diffusion-only baselines. The reported numbers on LCDP (24.09 dB PSNR, 0.8627 SSIM, 0.0905 LPIPS) and on MSEC/SICE averages support the claim that DPEC achieves state-of-the-art performance.

Load-bearing premise

The method assumes that a U-Net fine-tuned at a single fixed noise timestep and fed a 64x64 downsampled input yields low-frequency content more useful than a simple interpolation baseline, and that fusing these features into a full-resolution regressor improves fidelity rather than introducing hallucinated structure.

Editorial extensions

If this is right

  • Single-step sampling makes diffusion-based exposure correction practical at 0.31 seconds per 512x512 image, competitive with regression methods and far faster than multi-step diffusion baselines.
  • Because the diffusion branch operates only on low-frequency content at 64x64, artifacts common in fine-tuned diffusion outputs (such as distorted facial features or textures) are largely avoided.
  • On LCDP, MSEC, and SICE, DPEC reports the best or near-best PSNR, SSIM, LPIPS, and competitive NIQE, including on unlabeled generalization sets such as LIME, DICM, and NPE.
  • Removing the VAE and text conditioning from Stable Diffusion reduces computational overhead while retaining the benefit of the generative prior.

Reading between the lines

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

  • A testable implication not explored in the paper is that the same single-step fine-tuning recipe could transfer to other restoration tasks, such as dehazing or deraining, where the input provides most of the structure and only low-frequency content is missing.
  • The paper's ablation suggests the gain of the diffusion prior over a regression-only model is modest in PSNR (0.22 dB on LCDP) but larger in LPIPS (0.0138); if that pattern generalizes, the practical value of the diffusion prior may be mainly perceptual, which is exactly where regression methods are weakest.
  • Because the diffusion branch is fed a 64x64 input, the method's ceiling may depend on how well the downsampling preserves the low-frequency content of extreme exposure regions; a multi-scale or adaptive downsampling scheme could test whether the bottleneck is information loss or the denoiser itself.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes DPEC, a two-stage exposure correction framework built on a pre-trained Stable Diffusion model. In the first stage, the denoising U-Net is fine-tuned with a v-prediction objective at a fixed maximal timestep t=T, and inference is performed in a single denoising step on 64x64 downsampled images. In the second stage, the fine-tuned U-Net is frozen and its multi-scale features are fused into a regression encoder-decoder through a proposed Joint Cross-Attention Module (JCAM), with a regression loss combining L1, perceptual, and SSIM terms. The method is evaluated on LCDP, MSEC, and SICE with PSNR/SSIM/LPIPS/NIQE, on unlabeled LIME/DICM/NPE with NIQE, and compared against a large set of existing methods including diffusion-based restoration baselines and CoTF. Ablations cover the fine-tuning strategy, the noise timestep, the diffusion-regression integration, and the JCAM module.

Significance. If the reported results are reproducible, DPEC makes a useful practical contribution: it shows that a generative diffusion prior can be adapted to exposure correction with a single-step low-resolution denoising pass, avoiding the multi-step sampling cost of typical diffusion-based restoration. The division of labor between low-frequency generation by the diffusion model and high-frequency refinement by the regression network is a sensible design, and the mathematical relations in Eqs. (1)-(4) are correct. The paper also ships a reasonable set of ablations that are internally consistent. The significance is moderated, however, by the lack of any statistical uncertainty quantification and by the fact that the paper's own tables do not support the abstract's claim of consistent state-of-the-art performance across all reported fidelity, perceptual, and no-reference metrics.

major comments (4)
  1. [Abstract and Section IV-B, Tables I-III] The abstract claims that DPEC 'consistently outperforms existing state-of-the-art methods' across fidelity, perceptual quality, and visual effects, but the paper's own tables contradict this. On MSEC, CoTF [60] achieves higher average PSNR (23.44 vs 23.29) and better NIQE (3.70 vs 3.74). CoTF also achieves better NIQE on SICE (3.13 vs 3.28), LCDP (3.2123 vs 3.3592), LIME (3.17 vs 3.25), and DICM (3.10 vs 3.22). The authors should either revise the claim to state specifically which metrics and datasets DPEC improves upon (e.g., SICE PSNR/SSIM/LPIPS, LCDP PSNR/SSIM/LPIPS, NPE NIQE) or provide statistical significance tests demonstrating that the reported differences are meaningful. As written, the headline claim is not supported.
  2. [Section IV-C.2, Table V] The noise-level ablation is confounded. The text states that 'the test input is pure Gaussian noise,' which is only consistent with training at t=1000. For t=500 and t=1, the training-time forward process produces y_t distributions that differ substantially from pure Gaussian noise, so the poor results at those settings can be attributed to a train-test mismatch rather than to the noise level itself. A valid ablation would train and test with matched noise schedules at each t, or otherwise vary the inference-time noise distribution together with the training choice. Without this, the conclusion that t=1000 is the best design choice is not established.
  3. [Section IV-C.3, Table VI] The key load-bearing evidence that diffusion priors improve the final model is the comparison in Table VI: the full model beats 'Only Regression' by 0.22 dB PSNR, 0.0048 SSIM, and 0.0138 LPIPS on LCDP. No error bars, standard deviations, or significance tests are reported anywhere in the paper, so this gain could plausibly lie within run-to-run variability. Given that the entire architecture is justified by this improvement, the authors should provide repeated-run statistics (e.g., multiple training seeds with mean and std, or paired significance tests) for at least this central ablation.
  4. [Section III-B.2 and Table IV] The claim that 'increasing the number of time steps provides almost no additional improvement' is not directly supported by the reported experiments. Table IV compares 'Multi-noise 20' (22.03 dB), 'Multi-noise 1' (21.58 dB), and 'Single-noise 1' (21.97 dB), but these differ in both training strategy and sampling steps, so the comparison does not isolate the effect of the number of inference steps for a fixed model. A controlled experiment that varies only the sampling-step count of the same trained model is needed to substantiate the single-step efficiency claim.
minor comments (5)
  1. [Equation (3)] The expectation in L_diff is written as E_{x,y,t,eps}, but in the proposed strategy t is fixed to T. Please rewrite the objective with t=T made explicit, or clarify that a degenerate distribution over t is intended.
  2. [Section IV-B.1] The description of how the MSEC dataset is split into 'Under' and 'Over' subsets would be easier to follow if the text clarified whether the 'Average' column in Table I is the mean of the two subset columns.
  3. [Section IV-C.1, Table IV] The text says the proposed fine-tuning strategy 'demonstrates better performance in single-step generation,' but the single-noise 1-step setting (21.97 dB) is slightly lower than the multi-noise 20-step setting (22.03 dB). 'Comparable' would be a more accurate descriptor.
  4. [Section IV-C.4, Table VIII] The JCAM ablation reports PSNR, SSIM, and LPIPS. Adding a no-reference metric such as NIQE, or inference-time overhead, would strengthen the claim that JCAM improves the overall quality 'across multiple performance metrics.'
  5. [General] There are several typos and formatting inconsistencies, for example 'V AE' instead of 'VAE' in Sections III-C and IV-C.3, and the phrase 'the effect of the noise level' could be more specific as 'the effect of the noise level on reconstruction quality.'

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the DPEC derivation is trained against paired ground truth and evaluated against external benchmarks; the only author self-citation (CoTF) is a comparison baseline, not a load-bearing premise.

full rationale

I walked the claimed derivation chain and found no step where a 'prediction' or 'first-principles result' reduces by construction to its inputs. The fine-tuning loss (Eq. 3) is a standard v-prediction objective with paired ground-truth targets y, and the single-step sampling formula (Eq. 4) is algebraically derived from the same v-prediction definition, not from the result it is used to support. The multi-scale integration and JCAM are trained with the regression loss (Eq. 13) on L1, perceptual, and SSIM terms against ground truth, and the ablations (Tables IV-VI, VIII) compare components against regression-only, single-scale, and alternative attention variants on the external LCDP benchmark. The fixed noise timestep t=1000 is an ablated design choice (Table V), not a fitted parameter later renamed as a prediction. No uniqueness theorem or ansatz is imported from the authors' prior work to forbid alternatives. The only self-citation with overlapping authors is CoTF [60], used as a state-of-the-art comparison baseline and as the source of the CAM baseline in Table VIII; this is not load-bearing because the central architecture is trained from paired data and its effect is assessed independently of that citation. The abstract's 'consistently outperforms' claim is in tension with Tables I-III (where CoTF wins MSEC PSNR/NIQE, SICE NIQE, and LIME/DICM NIQE), but that is a correctness and evidence-strength concern, not circularity, and is outside the scope of this pass.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The method inherits the Stable Diffusion prior as a domain assumption rather than deriving it. The only hand-tuned values are t=1000, alpha=0.1, beta=0.5, and the 64x64 diffusion resolution. No new physical entities or postulates are introduced.

free parameters (3)
  • Diffusion noise timestep t = 1000
    Chosen by ablation on LCDP (Table V); the paper justifies it by matching test-time pure Gaussian noise under Eq. (4). This directly controls the denoising target.
  • Regression loss weights alpha and beta = alpha=0.1, beta=0.5
    Set by hand in Eq. (13) to balance L1, perceptual, and SSIM losses; no sensitivity analysis is provided.
  • Diffusion branch input resolution = 64x64
    Fixed to 64x64 to match Stable Diffusion v2-base and to define the low-frequency/high-frequency split; chosen by design rather than by analysis.
assumptions (4)
  • standard math DDPM v-prediction algebra in Eqs. (1)-(4): y_t = sqrt(alpha_bar_t) y0 + sqrt(1-alpha_bar_t) epsilon, v = sqrt(alpha_bar_t) epsilon - sqrt(1-alpha_bar_t) y0, and single-step reconstruction y0 = sqrt(alpha_bar_t) y_t - sqrt(1-alpha_bar_t) v.
    Needed to justify single-step sampling; this is standard DDPM/DDIM math, not proved in the paper.
  • domain assumption Stable Diffusion v2-base trained on LAION-5B encodes a useful general image prior that survives fine-tuning with only 100K iterations at 64x64 resolution.
    The central utility of the diffusion branch rests on this inheritance from [67]; the paper does not independently validate the prior beyond its own results.
  • domain assumption Downsampling to 64x64 leaves low-frequency content and removes high-frequency details, so the diffusion branch can improve perception without introducing artifacts.
    Sections III-B and III-C rest on this low-frequency/high-frequency split, but no frequency-domain analysis is provided to verify it.
  • domain assumption A fixed noise timestep T=1000 with single-step DDIM is sufficient for exposure correction because the conditioning image already carries pixel structure.
    Section III-B.2 relies on this; only ablations in Tables IV-V support it, and no proof is given that one step suffices across all exposure ranges.

how reviews work

0 comments
Cite this review

Pith. "Pith review of High-Quality Exposure Correction with Diffusion-Based Image Generation Priors." pith.science (2026). https://pith.science/paper/CZGQKS67

@misc{pith2026260808720,
  author       = {Pith},
  title        = {Pith review of: High-Quality Exposure Correction with Diffusion-Based Image Generation Priors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CZGQKS67}},
  note         = {Machine review of arXiv:2608.08720}
}
read the original abstract

Although most existing exposure correction methods achieve high fidelity, they often place excessive focus on overall pixel-wise accuracy, making it challenging to effectively model extreme exposure regions, which results in suboptimal perceptual quality. Recently, diffusion models have received significant attention due to their remarkable performance in the realm of image generation. However, their successful application to exposure correction remains a challenging and open question. The key challenge lies in generating accurate image structures and maintaining high image fidelity during stochastic diffusion processes. In this paper, we propose DPEC (Diffusion Prior-based Exposure Correction), a novel framework for image exposure correction that utilizes diffusion-based image generation priors encapsulated in pre-trained large-scale diffusion models. Specifically, we first propose an efficient fine-tuning strategy to derive an exposure corrector from pre-trained models, enabling the generation of enhanced images in a single-step denoising process. Moreover, we seamlessly combine the strengths of diffusion models and regression models, and design a joint cross-attention module to integrate multi-scale diffusion prior features, thereby effectively preserving high-frequency details and minimizing random artifacts. The diffusion model focuses on dealing with low-frequency content rather than all the intricate texture details. The experimental results demonstrate that the proposed DPEC method consistently outperforms existing state-of-the-art methods on multiple exposure correction datasets, whether in terms of fidelity, perceptual quality, or visual effects.

Figures

Figures reproduced from arXiv: 2608.08720 by the authors.

Figure 1
Figure 1. Comparison of visual effects. The regression-based methods, including (b), (c), and (d), may struggle to recover extreme exposure regions. Figure (e) [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The diagram illustrating the training and inference process of the fine-tuning strategy. Starting from the pre-trained stable diffusion model, we achieve [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Our DPEC overall framework: (a) Diffusion model: Perform single-step denoising in the low-resolution space. (b) Regression model: Use a hierarchical [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (10 more)
Figure 3
Figure 3. Figure 3: The goal is to effectively fuse the diffusion prior feature [PITH_FULL_IMAGE:figures/full_fig_p006_3.png]
Figure 4
Figure 4. Figure 4: Visual comparison with state-of-the-art methods on overexposed images from the SICE dataset. Best viewed by zooming in. [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Visual comparison with state-of-the-art methods on underexposed images from the SICE dataset. Best viewed by zooming in. [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 11
Figure 11. Figure 11: In Fig. 9, focusing the facial region, it is evident [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]
Figure 6
Figure 6. Figure 6: Visual comparison with state-of-the-art methods on overexposed images from the MSEC dataset. Best viewed by zooming in. [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Visual comparison with state-of-the-art methods on the unsupervised [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Ablation analysis on the diffusion and regression models. The output [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: Visual comparison with state-of-the-art methods on non-uniform illumination images from the LCDP dataset. Best viewed by zooming in. [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 11
Figure 11. Figure 11: Visual comparison with state-of-the-art methods on non-uniform illumination images from the LCDP dataset. Best viewed by zooming in. [PITH_FULL_IMAGE:figures/full_fig_p013_11.png]
Figure 12
Figure 12. Figure 12: Failure Cases. Our approach fails to correct large areas of overexpo [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

85 extracted references · 53 canonical work pages

  1. [60]

    Real- time exposure correction via collaborative transformations and adaptive sampling,

    Z. Li, F. Zhang, M. Cao, J. Zhang, Y . Shao, Y . Wang, and N. Sang, “Real- time exposure correction via collaborative transformations and adaptive sampling,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 2984–2994

    work page 2024

  2. [1]

    Planning- oriented autonomous driving,

    Y . Hu, J. Yang, L. Chen, K. Li, C. Sima, X. Zhu, S. Chai, S. Du, T. Lin, W. Wang, L. Lu, X. Jia, Q. Liu, J. Dai, Y . Qiao, and H. Li, “Planning- oriented autonomous driving,” inProceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, 2023

    2023

  3. [2]

    Iterative proposal refinement for weakly-supervised video grounding,

    M. Cao, F. Wei, C. Xu, X. Geng, L. Chen, C. Zhang, Y . Zou, T. Shen, and D. Jiang, “Iterative proposal refinement for weakly-supervised video grounding,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 6524–6534

    2023

  4. [3]

    Learning multi- scale photo exposure correction,

    M. Afifi, K. G. Derpanis, B. Ommer, and M. S. Brown, “Learning multi- scale photo exposure correction,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2021, pp. 9157–9167

    2021

  5. [4]

    Local color distributions prior for image enhancement,

    H. Wang, K. Xu, and R. W. Lau, “Local color distributions prior for image enhancement,” inProceedings of the European Conference on Computer Vision. Springer, 2022, pp. 343–359

    2022

  6. [5]

    Learning a deep single image contrast enhancer from multi-exposure images,

    J. Cai, S. Gu, and L. Zhang, “Learning a deep single image contrast enhancer from multi-exposure images,”IEEE Transactions on Image Processing, vol. 27, no. 4, pp. 2049–2062, 2018

    2018

  7. [6]

    Exploiting diffusion prior for real-world image super-resolution,

    J. Wang, Z. Yue, S. Zhou, K. C. Chan, and C. C. Loy, “Exploiting diffusion prior for real-world image super-resolution,”International Journal of Computer Vision, 2024. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 15 TABLE VII COMPARISON OF INFERENCE TIMES FOR DIFFERENT METHODS. ALL METHODS WERE TESTED ON512×512RESOLUTION IMAGES,WITH EACH METH...

    work page 2024

  8. [7]

    Pixel-aware stable dif- fusion for realistic image super-resolution and personalized stylization,

    T. Yang, R. Wu, P. Ren, X. Xie, and L. Zhang, “Pixel-aware stable dif- fusion for realistic image super-resolution and personalized stylization,” arXiv preprint arXiv:2308.14469, 2024

    arXiv 2024

Show all 85 references
  1. [8]

    Adaptive histogram equalization and its variations,

    S. M. Pizer, E. P. Amburn, J. D. Austin, R. Cromartie, A. Geselowitz, T. Greer, B. ter Haar Romeny, J. B. Zimmerman, and K. Zuiderveld, “Adaptive histogram equalization and its variations,”Computer vision, graphics, and image processing, 1987

    1987

  2. [9]

    Contrast limited adaptive histogram equalization,

    K. Zuiderveld, “Contrast limited adaptive histogram equalization,” Graphics gems, 1994

    1994

  3. [10]

    Adaptive image contrast enhancement using generalizations of histogram equalization,

    J. Stark, “Adaptive image contrast enhancement using generalizations of histogram equalization,”IEEE Transactions on Image Processing, vol. 9, no. 5, pp. 889–896, 2000

    2000

  4. [11]

    Video enhancement using per-pixel virtual exposures,

    E. P. Bennett and L. McMillan, “Video enhancement using per-pixel virtual exposures,”ACM Transactions on Graphics (TOG), vol. 24, no. 3, pp. 845–852, 2005

    2005

  5. [12]

    Automatic exposure correction of consumer photographs,

    L. Yuan and J. Sun, “Automatic exposure correction of consumer photographs,” inProceedings of the European Conference on Computer Vision. Springer, 2012, pp. 771–785

    2012

  6. [13]

    Properties and performance of a center/surround retinex,

    D. Jobson, Z. Rahman, and G. Woodell, “Properties and performance of a center/surround retinex,”IEEE Transactions on Image Processing, vol. 6, no. 3, pp. 451–462, 1997

    1997

  7. [14]

    Naturalness preserved enhancement algorithm for non-uniform illumination images,

    S. Wang, J. Zheng, H.-M. Hu, and B. Li, “Naturalness preserved enhancement algorithm for non-uniform illumination images,”IEEE Transactions on Image Processing, vol. 22, no. 9, pp. 3538–3548, 2013

    2013

  8. [15]

    Structure-revealing low- light image enhancement via robust retinex model,

    M. Li, J. Liu, W. Yang, X. Sun, and Z. Guo, “Structure-revealing low- light image enhancement via robust retinex model,”IEEE Transactions on Image Processing, vol. 27, no. 6, pp. 2828–2841, 2018

    2018

  9. [16]

    A weighted vari- ational model for simultaneous reflectance and illumination estimation,

    X. Fu, D. Zeng, Y . Huang, X.-P. Zhang, and X. Ding, “A weighted vari- ational model for simultaneous reflectance and illumination estimation,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 2782–2790

    2016

  10. [17]

    High-quality exposure correction of underexposed photos,

    Q. Zhang, G. Yuan, C. Xiao, L. Zhu, and W.-S. Zheng, “High-quality exposure correction of underexposed photos,” inProceedings of the 26th ACM international conference on Multimedia, 2018, pp. 582–590

    2018

  11. [18]

    Dual illumination estimation for robust exposure correction,

    Q. Zhang, Y . Nie, and W.-S. Zheng, “Dual illumination estimation for robust exposure correction,” inComputer graphics forum, vol. 38, no. 7. Wiley Online Library, 2019, pp. 243–252

    2019

  12. [19]

    Zero-shot restoration of underexposed images via robust retinex decomposition,

    A. Zhu, L. Zhang, Y . Shen, Y . Ma, S. Zhao, and Y . Zhou, “Zero-shot restoration of underexposed images via robust retinex decomposition,” inProceedings of the IEEE International Conference on Multimedia and Expo, 2020, pp. 1–6

    2020

  13. [20]

    Low-light image enhance- ment via self-reinforced retinex projection model,

    L. Ma, R. Liu, Y . Wang, X. Fan, and Z. Luo, “Low-light image enhance- ment via self-reinforced retinex projection model,”IEEE Transactions on Multimedia, pp. 1–1, 2022

    2022

  14. [21]

    Self-supervised low-light image enhancement using discrepant untrained network priors,

    J. Liang, Y . Xu, Y . Quan, B. Shi, and H. Ji, “Self-supervised low-light image enhancement using discrepant untrained network priors,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 11, pp. 7332–7345, 2022

    2022

  15. [22]

    Learning a simple low-light image enhancer from paired low-light instances,

    Z. Fu, Y . Yang, X. Tu, Y . Huang, X. Ding, and K.-K. Ma, “Learning a simple low-light image enhancer from paired low-light instances,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2023, pp. 22 252–22 261

    2023

  16. [23]

    Rethinking affine transform for efficient image enhancement: A color space perspective,

    D. Li and S. Rahardja, “Rethinking affine transform for efficient image enhancement: A color space perspective,”IEEE Transactions on Multi- media, pp. 1–12, 2024

    2024

  17. [24]

    Pixel-learnable 3dlut with saturation-aware compensation for image enhancement,

    J. Liu, Q. Li, X. Min, Y . Su, G. Zhai, and X. Yang, “Pixel-learnable 3dlut with saturation-aware compensation for image enhancement,”IEEE Transactions on Multimedia, vol. 26, pp. 11 219–11 231, 2024

    2024

  18. [25]

    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

    2018 arXiv

  19. [26]

    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

    2021

  20. [27]

    Kindling the darkness: A practical low-light image enhancer,

    Y . Zhang, J. Zhang, and X. Guo, “Kindling the darkness: A practical low-light image enhancer,” inProceedings of the 27th ACM Interna- tional Conference on Multimedia, 2019, pp. 1632–1640

    2019

  21. [28]

    Beyond brightening low-light images,

    Y . Zhang, X. Guo, J. Ma, W. Liu, and J. Zhang, “Beyond brightening low-light images,”International Journal of Computer Vision, vol. 129, pp. 1013–1037, 2021

    2021

  22. [29]

    Retinex-inspired unrolling with cooperative prior architecture search for low-light image enhancement,

    R. Liu, L. Ma, J. Zhang, X. Fan, and Z. Luo, “Retinex-inspired unrolling with cooperative prior architecture search for low-light image enhancement,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2021, pp. 10 556–10 565

    2021

  23. [30]

    Uretinex- net: Retinex-based deep unfolding network for low-light image enhance- ment,

    W. Wu, J. Weng, P. Zhang, X. Wang, W. Yang, and J. Jiang, “Uretinex- net: Retinex-based deep unfolding network for low-light image enhance- ment,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2022, pp. 5901–5910

    2022

  24. [31]

    Deep bilateral learning for real-time image enhancement,

    M. Gharbi, J. Chen, J. T. Barron, S. W. Hasinoff, and F. Durand, “Deep bilateral learning for real-time image enhancement,”ACM Transactions on Graphics (TOG), vol. 36, no. 4, pp. 1–12, 2017

    2017

  25. [32]

    Underexposed photo enhancement using deep illumination estimation,

    R. Wang, Q. Zhang, C.-W. Fu, X. Shen, W.-S. Zheng, and J. Jia, “Underexposed photo enhancement using deep illumination estimation,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2019, pp. 6842–6850

    2019

  26. [33]

    Toward fast, flexible, and robust low-light image enhancement,

    L. Ma, T. Ma, R. Liu, X. Fan, and Z. Luo, “Toward fast, flexible, and robust low-light image enhancement,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2022, pp. 5637–5646

    2022

  27. [34]

    Zero- reference deep curve estimation for low-light image enhancement,

    C. Guo, C. Li, J. Guo, C. C. Loy, J. Hou, S. Kwong, and R. Cong, “Zero- reference deep curve estimation for low-light image enhancement,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2020, pp. 1777–1786

    2020

  28. [35]

    Learning to enhance low-light image via zero-reference deep curve estimation,

    C. Li, C. Guo, and C. C. Loy, “Learning to enhance low-light image via zero-reference deep curve estimation,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 8, pp. 4225–4238, 2022

    2022

  29. [36]

    Luminance-aware pyramid network for low-light image enhancement,

    J. Li, J. Li, F. Fang, F. Li, and G. Zhang, “Luminance-aware pyramid network for low-light image enhancement,”IEEE Transactions on Multimedia, vol. 23, pp. 3153–3165, 2021

    2021

  30. [37]

    Dslr: Deep stacked laplacian restorer for low- light image enhancement,

    S. Lim and W. Kim, “Dslr: Deep stacked laplacian restorer for low- light image enhancement,”IEEE Transactions on Multimedia, vol. 23, pp. 4272–4284, 2021

    2021

  31. [38]

    Illumination guided attentive wavelet network for low-light image enhancement,

    J. Xu, M. Yuan, D.-M. Yan, and T. Wu, “Illumination guided attentive wavelet network for low-light image enhancement,”IEEE Transactions on Multimedia, pp. 1–14, 2022

    2022

  32. [39]

    Half wavelet attention on m- net+ for low-light image enhancement,

    C.-M. Fan, T.-J. Liu, and K.-H. Liu, “Half wavelet attention on m- net+ for low-light image enhancement,” inProceedings of the IEEE International Conference on Image Processing, 2022, pp. 3878–3882

    2022

  33. [40]

    Low-light image enhancement with multi-scale attention and frequency-domain optimization,

    Z. He, W. Ran, S. Liu, K. Li, J. Lu, C. Xie, Y . Liu, and H. Lu, “Low-light image enhancement with multi-scale attention and frequency-domain optimization,”IEEE Transactions on Circuits and Systems for Video Technology, pp. 1–1, 2023

    2023

  34. [41]

    Enlightengan: Deep light enhancement without paired supervision,

    Y . Jiang, X. Gong, D. Liu, Y . Cheng, C. Fang, X. Shen, J. Yang, P. Zhou, and Z. Wang, “Enlightengan: Deep light enhancement without paired supervision,”IEEE Transactions on Image Processing, vol. 30, pp. 2340–2349, 2021

    2021

  35. [42]

    Towards unsupervised deep image enhancement with generative adversarial network,

    Z. Ni, W. Yang, S. Wang, L. Ma, and S. Kwong, “Towards unsupervised deep image enhancement with generative adversarial network,”IEEE Transactions on Image Processing, vol. 29, pp. 9140–9151, 2020. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 16

    2020

  36. [43]

    Deep photo enhancer: Unpaired learning for image enhancement from photographs with gans,

    Y .-S. Chen, Y .-C. Wang, M.-H. Kao, and Y .-Y . Chuang, “Deep photo enhancer: Unpaired learning for image enhancement from photographs with gans,” in2018 IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 6306–6314

    2018

  37. [44]

    Cross-image disentan- glement for low-light enhancement in real world,

    L. Guo, R. Wan, W. Yang, A. Kot, and B. Wen, “Cross-image disentan- glement for low-light enhancement in real world,”IEEE Transactions on Circuits and Systems for Video Technology, pp. 1–1, 2023

    2023

  38. [45]

    Band representation- based semi-supervised low-light image enhancement: Bridging the gap between signal fidelity and perceptual quality,

    W. Yang, S. Wang, Y . Fang, Y . Wang, and J. Liu, “Band representation- based semi-supervised low-light image enhancement: Bridging the gap between signal fidelity and perceptual quality,”IEEE Transactions on Image Processing, vol. 30, pp. 3461–3473, 2021

    2021

  39. [46]

    Low-light image enhancement with normalizing flow,

    Y . Wang, R. Wan, W. Yang, H. Li, L.-P. Chau, and A. Kot, “Low-light image enhancement with normalizing flow,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 36, no. 3, 2022, pp. 2604– 2612

    2022

  40. [47]

    Snr-aware low-light image enhancement,

    X. Xu, R. Wang, C.-W. Fu, and J. Jia, “Snr-aware low-light image enhancement,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2022, pp. 17 693–17 703

    2022

  41. [48]

    Ultra-high- definition low-light image enhancement: A benchmark and transformer- based method,

    T. Wang, K. Zhang, T. Shen, W. Luo, B. Stenger, and T. Lu, “Ultra-high- definition low-light image enhancement: A benchmark and transformer- based method,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 3, 2023, pp. 2654–2662

    2023

  42. [49]

    Bilateral interaction for local-global collaborative perception in low-light image enhancement,

    R. Xu, Y . Li, Y . Niu, H. Xu, Y . Chen, and T. Zhao, “Bilateral interaction for local-global collaborative perception in low-light image enhancement,”IEEE Transactions on Multimedia, vol. 26, pp. 10 792– 10 804, 2024

    2024

  43. [50]

    Low-light image enhancement with sam- based structure priors and guidance,

    G. Li, B. Zhao, and X. Li, “Low-light image enhancement with sam- based structure priors and guidance,”IEEE Transactions on Multimedia, vol. 26, pp. 10 854–10 866, 2024

    2024

  44. [51]

    Learning semantic-aware knowledge guidance for low-light image enhancement,

    Y . Wu, C. Pan, G. Wang, Y . Yang, J. Wei, C. Li, and H. T. Shen, “Learning semantic-aware knowledge guidance for low-light image enhancement,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2023, pp. 1662–1671

    2023

  45. [52]

    From fidelity to perceptual quality: A semi-supervised approach for low-light image enhancement,

    W. Yang, S. Wang, Y . Fang, Y . Wang, and J. Liu, “From fidelity to perceptual quality: A semi-supervised approach for low-light image enhancement,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2020, pp. 3060–3069

    2020

  46. [53]

    Exposure normalization and compensation for multiple-exposure cor- rection,

    J. Huang, Y . Liu, X. Fu, M. Zhou, Y . Wang, F. Zhao, and Z. Xiong, “Exposure normalization and compensation for multiple-exposure cor- rection,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2022, pp. 6043–6052

    2022

  47. [54]

    Exposure- consistency representation learning for exposure correction,

    J. Huang, M. Zhou, Y . Liu, M. Yao, F. Zhao, and Z. Xiong, “Exposure- consistency representation learning for exposure correction,” inProceed- ings of the 30th ACM International Conference on Multimedia, 2022, pp. 6309–6317

    2022

  48. [55]

    Deep fourier-based exposure correction network with spatial- frequency interaction,

    J. Huang, Y . Liu, F. Zhao, K. Yan, J. Zhang, Y . Huang, M. Zhou, and Z. Xiong, “Deep fourier-based exposure correction network with spatial- frequency interaction,” inProceedings of the European Conference on Computer Vision. Springer, 2022, pp. 163–180

    2022

  49. [56]

    Decoupling-and- aggregating for image exposure correction,

    Y . Wang, L. Peng, L. Li, Y . Cao, and Z.-J. Zha, “Decoupling-and- aggregating for image exposure correction,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2023, pp. 18 115–18 124

    2023

  50. [57]

    Cudi: Curve distillation for efficient and controllable exposure adjustment,

    C. Li, C. Guo, R. Feng, S. Zhou, and C. C. Loy, “Cudi: Curve distillation for efficient and controllable exposure adjustment,”arXiv preprint arXiv:2207.14273, 2022

    2022 arXiv

  51. [58]

    Learning sample relationship for exposure correction,

    J. Huang, F. Zhao, M. Zhou, J. Xiao, N. Zheng, K. Zheng, and Z. Xiong, “Learning sample relationship for exposure correction,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2023, pp. 9904–9913

    2023

  52. [59]

    Iterative prompt learning for unsupervised backlit image enhancement,

    Z. Liang, C. Li, S. Zhou, R. Feng, and C. C. Loy, “Iterative prompt learning for unsupervised backlit image enhancement,” inProceedings of the IEEE International Conference on Computer Vision, 2023, pp. 8094–8103

    2023

  53. [61]

    Osmamba: Omnidi- rectional spectral mamba with dual-domain prior generator for exposure correction,

    G. Li, B. Chen, C. Zhao, L. Zhang, and J. Zhang, “Osmamba: Omnidi- rectional spectral mamba with dual-domain prior generator for exposure correction,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2025, pp. 7480–7490

    2025

  54. [62]

    Clip-restorex: Restore image structure and perception in exposure correction,

    X. Huang, Q. Zhang, J.-F. Hu, and W.-S. Zheng, “Clip-restorex: Restore image structure and perception in exposure correction,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 4. AAAI, 2025, pp. 3760–3768

    2025

  55. [63]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems, vol. 33, pp. 6840– 6851, 2020

    2020

  56. [64]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” arXiv preprint arXiv:2010.02502, 2020

    2010 arXiv

  57. [65]

    Diffusion models beat gans on image synthesis,

    P. Dhariwal and A. Nichol, “Diffusion models beat gans on image synthesis,”Advances in neural information processing systems, vol. 34, pp. 8780–8794, 2021

    2021

  58. [66]

    Classifier-free diffusion guidance,

    J. Ho and T. Salimans, “Classifier-free diffusion guidance,”arXiv preprint arXiv:2207.12598, 2022

    2022 arXiv

  59. [67]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 10 684–10 695

    2022

  60. [68]

    Diffbir: Toward blind image restoration with generative diffusion prior,

    X. Lin, J. He, Z. Chen, Z. Lyu, B. Dai, F. Yu, Y . Qiao, W. Ouyang, and C. Dong, “Diffbir: Toward blind image restoration with generative diffusion prior,” inEuropean conference on computer vision. Springer, 2024, pp. 430–448

    2024

  61. [69]

    Adding conditional control to text-to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text-to-image diffusion models,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 3836–3847

    2023

  62. [70]

    Exposurediffusion: Learning to expose for low-light image enhancement,

    Y . Wang, Y . Yu, W. Yang, L. Guo, L.-P. Chau, A. C. Kot, and B. Wen, “Exposurediffusion: Learning to expose for low-light image enhancement,” inIEEE/CVF International Conference on Computer Vision (ICCV). IEEE, 2023, pp. 12 438–12 448

    2023

  63. [71]

    Diff-retinex: Rethinking low-light image enhancement with a generative diffusion model,

    X. Yi, H. Xu, H. Zhang, L. Tang, and J. Ma, “Diff-retinex: Rethinking low-light image enhancement with a generative diffusion model,” in Proceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 12 302–12 311

    2023

  64. [72]

    Dimensionx: Create any 3d and 4d scenes from a single image with controllable video diffusion,

    W. Sun, S. Chen, F. Liu, Z. Chen, Y . Duan, J. Zhang, and Y . Wang, “Dimensionx: Create any 3d and 4d scenes from a single image with controllable video diffusion,”arXiv preprint arXiv:2411.04928, 2024

    2024 arXiv

  65. [73]

    Videoscene: Distilling video diffusion model to generate 3d scenes in one step,

    H. Wang, F. Liu, J. Chi, and Y . Duan, “Videoscene: Distilling video diffusion model to generate 3d scenes in one step,” in2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2025, pp. 16 475–16 485

    2025

  66. [74]

    Reconx: Reconstruct any scene from sparse views with video diffusion model,

    F. Liu, W. Sun, H. Wang, Y . Wang, H. Sun, J. Ye, J. Zhang, and Y . Duan, “Reconx: Reconstruct any scene from sparse views with video diffusion model,”arXiv preprint arXiv:2408.16767, 2024

    2024 arXiv

  67. [75]

    Physics3d: Learning physical properties of 3d gaussians via video diffusion,

    F. Liu, H. Wang, S. Yao, S. Zhang, J. Zhou, and Y . Duan, “Physics3d: Learning physical properties of 3d gaussians via video diffusion,”arXiv preprint arXiv:2406.04338, 2024

    2024 arXiv

  68. [76]

    Fourier priors-guided diffusion for zero-shot joint low-light enhance- ment and deblurring,

    X. Lv, S. Zhang, C. Wang, Y . Zheng, B. Zhong, C. Li, and L. Nie, “Fourier priors-guided diffusion for zero-shot joint low-light enhance- ment and deblurring,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 25 378–25 388

    2024

  69. [77]

    Zero-shot low light image enhancement with diffusion prior,

    J. Cho, S. Aghajanzadeh, Z. Zhu, and D. Forsyth, “Zero-shot low light image enhancement with diffusion prior,”arXiv preprint arXiv:2412.13401, 2024

    2024 arXiv

  70. [78]

    Sagiri: Low dynamic range image enhancement with generative diffu- sion prior,

    B. Li, S. Ma, Y . Zeng, X. Xu, Y . Fang, Z. Zhang, J. Wang, and K. Chen, “Sagiri: Low dynamic range image enhancement with generative diffu- sion prior,”arXiv preprint arXiv:2406.09389, 2024

    2024 arXiv

  71. [79]

    Lime: Low-light image enhancement via illumination map estimation,

    X. Guo, Y . Li, and H. Ling, “Lime: Low-light image enhancement via illumination map estimation,”IEEE Transactions on Image Processing, vol. 26, no. 2, pp. 982–993, 2017

    2017

  72. [80]

    Learning to see in the dark,

    C. Chen, Q. Chen, J. Xu, and V . Koltun, “Learning to see in the dark,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 3291–3300

    2018

  73. [81]

    Adam: A method for stochastic optimization,

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

    2014 arXiv

  74. [82]

    Sgdr: Stochastic gradient descent with warm restarts,

    I. Loshchilov and F. Hutter, “Sgdr: Stochastic gradient descent with warm restarts,” inInternational Conference on Learning Representa- tions, 2016

    2016

  75. [83]

    Image quality assessment: from error visibility to structural similarity,

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

    2004

  76. [84]

    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

    2018

  77. [85]

    Making a “completely blind

    A. Mittal, R. Soundararajan, and A. C. Bovik, “Making a “completely blind” image quality analyzer,”IEEE Signal processing letters, vol. 20, no. 3, pp. 209–212, 2012

    2012

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.