Pith. sign in

REVIEW 5 major objections 6 minor 61 references

VIIS: Visible and Infrared Information Synthesis for Severe Low-light Image Enhancement

T0 review · 5 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper introduces a new task, visible and infrared information synthesis (VIIS), in which a severely dark visible image and its paired infrared image are jointly enhanced and fused into a single high-quality color image.

desk verdict A novel task and a plausible diffusion method, but the circular fidelity test and weak no-reference metrics don't support the state-of-the-art claim. read the letter →

arxiv 2412.13655 v2 pith:DRXKF3CU submitted 2024-12-18 cs.CV

Chen Zhao , Mengyuan Yu , Fan Yang , Peiguang Jing This is my paper · ORCID
classification cs.CV
keywords low-lightimageenhancementinfraredandvisiblefusiondiffusionmodelinformationsynthesissparsecross-attentionpretexttaskmulti-modalconditioningrestoration
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 proposes a new task, visible and infrared information synthesis (VIIS), in which a pair of severely dark visible and infrared images is turned into a single bright, colored, detail-rich image. Existing low-light enhancement methods fail where visible information is completely absent, and existing fusion methods only combine modalities without enhancing them. The authors argue that doing both at once—enhancing each modality and fusing them—solves the ill-posed problem, because the infrared image supplies the missing structure while the visible image supplies color and texture. To make training possible without ground truth, they design an augmentation-based pretext task that degrades daytime images into pseudo-low-light pairs. The paper reports that their diffusion-based model outperforms state-of-the-art low-light enhancement, infrared colorization, and fusion methods on standard datasets.

What carries the argument

The key machinery is the sparse attention-based dual-modalities residual (SADMR) conditioning mechanism inside a latent diffusion model (LDM). The visible and infrared images are encoded by a pre-trained VAE, concatenated with the noise, and also encoded into multi-scale features that are injected in parallel into the encoder and decoder of the denoising U-Net. In place of simple addition or concatenation, a sparse cross-attention module based on deformable attention lets each element of the intermediate features attend only to sampled points around its reference location in both modality feature maps, enabling adaptive, iterative cross-modal interaction during denoising.

What would settle it

A direct test is to collect a small set of real visible–infrared pairs captured in genuinely severe low light, with human-verified ground-truth content for objects hidden in darkness, and compare the model's output against the original visible image and the infrared image. If the model does not reliably reveal objects that are invisible in the visible input but present in the infrared input—or if its outputs on real data are markedly worse than on pseudo-low-light images from the same daytime scenes—the claim that the pretext task transfers would be falsified.

Watch

Extended reading notes

Core claim

The central claim is that simultaneous intra-modal enhancement and inter-modal fusion of visible and infrared images is achievable in a single diffusion-based model, and that this joint formulation outperforms both singular-modality enhancement and conventional fusion. The model, VIIS, is trained with an information synthesis pretext task: daytime visible images are degraded by gamma darkening, contrast reduction, and Gaussian and Poisson noise to create pseudo-low-light inputs, and the model must recover the original image using the corresponding infrared image as guidance. During denoising, a sparse attention-based dual-modalities residual (SADMR) conditioning mechanism lets intermediate features of the U-Net attend to each modality's features iteratively. On the MSRS and KAIST-MS datasets, the authors find the model produces higher perceptual quality and better information recovery than ten existing methods and four newly constructed pipeline baselines.

Load-bearing premise

The load-bearing premise is that the synthetic degradation used in the pretext task—intense gamma darkening, contrast reduction, and Gaussian and Poisson noise—sufficiently mimics real severe low-light conditions, so that a model trained on pseudo-low-light daytime images transfers to real nighttime and subterranean imagery.

Editorial extensions

If this is right

  • A model trained on pseudo-low-light pairs can restore regions where the visible image carries no information, by borrowing structure from infrared and color from visible cues.
  • The same diffusion backbone with the SADMR conditioning mechanism can be applied to other multi-modal synthesis tasks that require iterative cross-modal interaction.
  • Because the pretext task requires no paired ground truth, the training procedure can be adapted to any aligned visible–infrared dataset with daytime images.
  • The reported results imply that joint enhancement and fusion is preferable to chaining separate enhancement, colorization, and fusion modules.

Reading between the lines

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

  • A natural extension the authors do not explore is whether the same pretext-task design transfers to other modality pairs, such as visible and depth or visible and event cameras, where one modality is robust to darkness.
  • The fidelity experiment only tests synthetic pseudo-low-light degradation; a real-world benchmark with manual annotations for objects hidden in darkness would be a stronger test of the transfer claim.
  • The sparse attention mechanism's sampling offsets could be visualized to check whether the model is genuinely using infrared structure in the darkest regions, which would make the fusion mechanism more interpretable.
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

5 major / 6 minor

Summary. The paper introduces a new task, visible and infrared information synthesis (VIIS), which aims to jointly enhance severely low-light visible images and colorize infrared images into a single high-quality output. To overcome the absence of ground truth, the authors propose an information synthesis pretext task (ISPT) based on gamma/contrast/noise augmentation, and a sparse attention-based dual-modalities residual (SADMR) conditioning mechanism for a latent diffusion model. Experiments on MSRS and KAIST-MS, including no-reference metric comparisons, ablations, a user study, and a synthetic fidelity test, are presented to support the claim that the method outperforms state-of-the-art low-light enhancement, infrared colorization, and visible-infrared fusion methods as well as newly constructed pipeline baselines.

Significance. If the claimed performance is substantiated, VIIS would be a useful contribution to severe low-light imaging, combining intra-modal enhancement and inter-modal fusion in a single diffusion-based model. The paper offers a novel task formulation, a publicly available codebase, and extensive ablations of the conditioning mechanism and the pretext task components. The methodological novelty (SADMR) is clearly described. However, the current evaluation does not convincingly establish the central claim: the fidelity experiment is circular, and the real-image comparison relies on no-reference metrics without statistical validation. The contribution is therefore promising but not yet demonstrated to the standard required for the claimed state-of-the-art results.

major comments (5)
  1. [Sec. 9, Table 8] The fidelity evaluation is circular: the synthetic test inputs are generated by the same ISPT augmentation (γ=6, α=0.5, λ=10, σ=5) used in training, and these parameters lie inside the training ranges given in Sec. 4.1 (γ∈[3,10], α∈[0.1,1], σ∈[0,10], λ∈[0,20]). Consequently the PSNR/SSIM gains (+3 dB and +0.07 in Table 8) may reflect the model's ability to invert its own training degradation rather than genuine generalization to real severe low-light conditions. The authors should provide non-circular evidence, e.g., testing with a different degradation model or with real paired low-light/high-light data.
  2. [Sec. 4.2, Tables 2 and 3] The real-image evaluation relies on no-reference metrics (NIQE, BRISQUE, SD, EN) computed on small, subjectively selected subsets (62 and 68 pairs) without error bars or significance tests. NIQE and BRISQUE are known to be imperfect proxies for perceptual quality and can be improved by increased contrast or colorfulness; SD and EN reward high contrast and high entropy, which may favor noisy outputs. This evidence is not sufficient to support the claim that the method outperforms baselines on real severe low-light images.
  3. [Sec. 8, Table 6] The user study uses only 20 images and 15 participants, and the voting percentages show large margins but no statistical analysis (e.g., confidence intervals or significance testing). The small sample size and the fact that most participants have computer vision knowledge limit the strength of the conclusions drawn from this study.
  4. [Sec. 4.1, Tables 1 and 7] The newly designed baselines (EMD, PSMD, RMD, PDMD, UMD) are pipelines that chain independently pretrained enhancement, colorization, and fusion networks without any joint training for the VIIS objective. Beating these baselines is weak evidence of superiority over methods that are actually optimized for the combined task; the comparison against individual task-specific SOTA methods, while informative, does not directly address the joint VIIS setting.
  5. [Sec. 5] The authors acknowledge that their pretext task may weaken visible-specific information (e.g., texture or color unique to the visible image), but this trade-off is not quantified. Given that the abstract claims perceptual quality superiority, a quantitative analysis of this failure mode (e.g., how often and where visible-only cues are lost) would help the reader assess the practical applicability of the method.
minor comments (6)
  1. [Sec. 4.3] The ablation table in the main text is labeled Table 4 but is referenced as 'Tab. 5' in the text; please fix the cross-references.
  2. [Sec. 9.1] The phrase 'data argumentation' should be 'data augmentation'.
  3. [Eq. (7)] The Gaussian noise term is written as N(0, σ) and the Poisson noise as P(λ); clarify whether σ is the standard deviation or variance, and specify the parameterization and domain of λ.
  4. [Fig. 2] The caption 'The overall of our model' should be 'Overview of our model'.
  5. [Sec. 4.1] Please state the sampling distribution for the augmentation parameters (e.g., uniform over the stated ranges) and whether the same random draw is used for both modalities; this is important for reproducibility.
  6. [General] The paper would benefit from reporting inference runtime and model complexity, especially since diffusion models are typically computationally expensive.

Circularity Check

1 steps flagged · score 6.0 of 10

The fidelity experiment (Sec. 9) is circular: test images are degraded by the same ISPT augmentation that defines the training objective, so the reported PSNR/SSIM gains measure inversion of the training pretext rather than fidelity on real severe low-light images.

  1. fitted input called prediction [Section 9 and Section 9.1 (also Section 3.4 and Section 4.1)]
    "we construct a synthetic dataset. Specifically, we conduct the fidelity experiment on KAIST-MS [17] dataset and employ the image augmentation method of the information synthesis pretext task (ISPT) to degrade clear high-quality images in the daytime testing set. ... the parameters of the image augmentation for the testing dataset are set as follows: the gamma parameter γ is configured at 6, the contrast factor α is set to 0.5, and the noise parameters are set as λ = 10, σ= 5."

    The model is trained on pseudo-low-light images produced by ISPT (gamma transform, contrast reduction, Gaussian/Poisson noise), with training ranges gamma [3,10], alpha [0.1,1], sigma [0,10], lambda [0,20] (Sec. 4.1). The fidelity test applies the same ISPT degradation to daytime images, using test parameters (gamma=6, alpha=0.5, lambda=10, sigma=5) that fall inside those training ranges. The task is therefore to invert the exact augmentation family the model was optimized to invert. The reported +3 PSNR / +0.07 SSIM margin (Tab. 8) quantifies how well the model learned the pretext, not how faithfully it restores real severe low-light scenes, whose degradation process is never shown to match ISPT. The real-image evaluation (Tabs.

full rationale

The paper's architectural contribution (diffusion model with SADMR conditioning) and the real-image comparisons are not circular: the model is trained on ISPT-degraded daytime images and compared against baselines on real MSRS and KAIST-MS night images using no-reference metrics. That evidence is independent but weak, since NIQE/BRISQUE/SD/EN do not establish fidelity and the real test set has no ground truth. The decisive fidelity evidence in Sec. 9 is circular: the test inputs are generated by the same gamma/contrast/noise augmentation that defines the ISPT training objective, and the test parameters lie within the training ranges. Consequently, the PSNR/SSIM advantage measures the model's ability to invert its own training augmentation, not its ability to restore authentic information under real severe low-light conditions. This is a fitted-input-called-prediction pattern: the degradation model is both the training target generator and the test oracle. The paper does not provide non-circular evidence that ISPT transfers to real severe low-light imagery. There is no load-bearing self-citation or uniqueness argument. Overall, the central SOTA claim is partially grounded, but a key quantitative pillar reduces by construction, giving a circularity score of 6.

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

The method rests on hand-chosen augmentation parameters that define the entire supervisory signal, and on the assumption that these synthetic degradations transfer to real severe low-light conditions. No new physical entities are introduced.

free parameters (5)
  • ISPT gamma range = [3, 10]
    Gamma transform exponent range used to darken day-time images; chosen by hand to simulate severe low light.
  • ISPT contrast factor range = [0.1, 1]
    Contrast adjustment factor range to emulate low contrast.
  • ISPT Gaussian noise sigma range = [0, 10]
    Standard deviation of Gaussian noise added to emulate electronic noise.
  • ISPT Poisson noise lambda range = [0, 20]
    Mean parameter of Poisson noise to emulate shot noise.
  • Fidelity test augmentation parameters = gamma=6, alpha=0.5, lambda=10, sigma=5
    Fixed parameters used to synthesize test inputs in the fidelity experiment; chosen by hand from the training ranges.
assumptions (4)
  • domain assumption Aligned visible-infrared pairs in MSRS and KAIST-MS are valid for training and evaluation
    The method assumes the two modalities are spatially aligned and that daytime infrared images are informative for nighttime-like degraded visible images.
  • ad hoc to paper Gamma transform, contrast reduction, and noise approximate severe low-light degradation
    The ISPT is defined by this degradation model; the fidelity evaluation uses the same model, making the reconstruction task close to the training objective.
  • domain assumption Pretrained VQGAN encoder/decoder from LDM generalizes to infrared image latents
    The model encodes both visible and infrared images with the same pretrained encoder, assuming the encoder represents infrared content adequately.
  • domain assumption Sparse deformable attention provides sufficient multi-modal interaction
    The design choice that deformable attention is effective for cross-modal conditioning; not proven independently.

how reviews work

0 comments
Cite this review

Pith. "Pith review of VIIS: Visible and Infrared Information Synthesis for Severe Low-light Image Enhancement." pith.science (2026). https://pith.science/paper/DRXKF3CU

@misc{pith2026241213655,
  author       = {Pith},
  title        = {Pith review of: VIIS: Visible and Infrared Information Synthesis for Severe Low-light Image Enhancement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DRXKF3CU}},
  note         = {Machine review of arXiv:2412.13655}
}
read the original abstract

Images captured in severe low-light circumstances often suffer from significant information absence. Existing singular modality image enhancement methods struggle to restore image regions lacking valid information. By leveraging light-impervious infrared images, visible and infrared image fusion methods have the potential to reveal information hidden in darkness. However, they primarily emphasize inter-modal complementation but neglect intra-modal enhancement, limiting the perceptual quality of output images. To address these limitations, we propose a novel task, dubbed visible and infrared information synthesis (VIIS), which aims to achieve both information enhancement and fusion of the two modalities. Given the difficulty in obtaining ground truth in the VIIS task, we design an information synthesis pretext task (ISPT) based on image augmentation. We employ a diffusion model as the framework and design a sparse attention-based dual-modalities residual (SADMR) conditioning mechanism to enhance information interaction between the two modalities. This mechanism enables features with prior knowledge from both modalities to adaptively and iteratively attend to each modality's information during the denoising process. Our extensive experiments demonstrate that our model qualitatively and quantitatively outperforms not only the state-of-the-art methods in relevant fields but also the newly designed baselines capable of both information enhancement and fusion. The code is available at https://github.com/Chenz418/VIIS.

Figures

Figures reproduced from arXiv: 2412.13655 by the authors.

Figure 1
Figure 1. An existing low-light enhancement method Enlight [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overall of our model. The information synthesis pretext task (ISPT) initially generates the pseudo-low-light visible images [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Qualitative comparison of ’01012N’ from MSRS dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison of ’01198N’ from MSRS dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison of ’set09 v000 00330’ from KAIST-MS dataset [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Qualitative comparison of ’set09 v000 00805’ from KAIST-MS dataset. evaluator (BRISQUE) ↓ [36] as metrics. EN measures the amount of information, SD assesses the contrast, while NIQE and BRISQUE evaluate the comprehensive percep￾tual quality. As shown in Tabs. 2 and 3,…
Figure 7
Figure 7. Figure 7: In this instance, although our method enhances the over [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: illustrates the data augmentation process of ISPT, where each component is systematically ablated to verify its effectiveness. Specifically, we first reduce the influence of the gamma transform and contrast adjustment operations separately. For the gamma transform, the…
Figure 9
Figure 9. Figure 9: Qualitative results on discussing components of ISPT. [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: Screenshot of the questionary interface from user study [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: Some instances in the testing dataset where the high [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]
Figure 12
Figure 12. Figure 12: Qualitative comparisons for fidelity experiment. The ground truth is shown on the left of Fig. [PITH_FULL_IMAGE:figures/full_fig_p011_12.png]
Figure 13
Figure 13. Figure 13: In SCAM, the intermediate feature F l e from the Unet￾based denoising network and visible/infrared features F l v, Fl i are initially embedded to get muti-head queries and values. Subse￾quently, each query element attends to elements surrounding its reference location…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

61 extracted references · 52 canonical work pages

  1. [1]

    Learning op- timized low-light image enhancement for edge vision tasks

    SM A Sharif, Azamat Myrzabekov, Nodirkhuja Khudjaev, Roman Tsoy, Seongwan Kim, and Jaeho Lee. Learning op- timized low-light image enhancement for edge vision tasks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6373–6383, 2024. 1, 2

    work page 2024

  2. [2]

    Optimizing observ- ing strategies for monitoring animals using drone-mounted thermal infrared cameras

    Claire Burke, Maisie Rashman, Serge Wich, Andy Symons, Cobus Theron, and Steve Longmore. Optimizing observ- ing strategies for monitoring animals using drone-mounted thermal infrared cameras. International Journal of Remote Sensing, 40(2):439–467, 2019. 1, 8

    work page 2019

  3. [3]

    Retinexformer: One-stage retinex- based transformer for low-light image enhancement

    Yuanhao Cai, Hao Bian, Jing Lin, Haoqian Wang, Radu Tim- ofte, and Yulun Zhang. Retinexformer: One-stage retinex- based transformer for low-light image enhancement. In Pro- ceedings of the IEEE International Conference on Computer Vision, pages 12504–12513, 2023. 1, 2, 5, 6, 7, 8, 10, 11

    work page 2023

  4. [4]

    Infrared thermal imaging in stud- ies of wild animals

    Justyna Cilulko, Paweł Janiszewski, Marek Bogdaszewski, and Eliza Szczygielska. Infrared thermal imaging in stud- ies of wild animals. European Journal of Wildlife Research, 59:17–23, 2013. 1, 8

    work page 2013

  5. [5]

    Diffusion models beat gans on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in Neural Informa- tion Processing Systems, pages 8780–8794, 2021. 3

    work page 2021

  6. [6]

    Taming transformers for high-resolution image synthesis

    Patrick Esser, Robin Rombach, and Bjorn Ommer. Taming transformers for high-resolution image synthesis. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 12873–12883, 2021. 3

    work page 2021

  7. [7]

    Le-gan: Unsupervised low-light image enhancement network using attention module and identity invariant loss

    Ying Fu, Yang Hong, Linwei Chen, and Shaodi You. Le-gan: Unsupervised low-light image enhancement network using attention module and identity invariant loss. Knowledge- Based Systems, 240:108010, 2022. 2

    work page 2022

  8. [8]

    Image fusion based on generative adversarial network consistent with percep- tion

    Yu Fu, Xiao-Jun Wu, and Tariq Durrani. Image fusion based on generative adversarial network consistent with percep- tion. Information Fusion, 72:110–125, 2021. 2

    work page 2021

Show all 61 references
  1. [9]

    Implicit diffusion models for continuous super-resolution

    Sicheng Gao, Xuhui Liu, Bohan Zeng, Sheng Xu, Yan- jing Li, Xiaoyan Luo, Jianzhuang Liu, Xiantong Zhen, and Baochang Zhang. Implicit diffusion models for continuous super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 10021–...

    2023

  2. [10]

    Practical markov chain monte carlo

    Charles J Geyer. Practical markov chain monte carlo. Statis- tical science, pages 473–483, 1992. 3

    1992

  3. [11]

    Denoising diffu- sion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffu- sion probabilistic models. Advances in Neural Information Processing Systems, pages 6840–6851, 2020. 2, 3

    2020

  4. [12]

    Cascaded diffusion models for high fidelity image generation

    Jonathan Ho, Chitwan Saharia, William Chan, David J Fleet, Mohammad Norouzi, and Tim Salimans. Cascaded diffusion models for high fidelity image generation. The Journal of Machine Learning Research, 23(1):2249–2281, 2022. 3

    2022

  5. [13]

    Classifier-free diffusion guidance

    Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance. In NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications, 2021. 3

    2021

  6. [14]

    Global structure-aware diffusion pro- cess for low-light image enhancement

    Jinhui Hou, Zhiyu Zhu, Junhui Hou, Hui Liu, Huanqiang Zeng, and Hui Yuan. Global structure-aware diffusion pro- cess for low-light image enhancement. Advances in Neural Information Processing Systems, 36, 2024. 1, 3

    2024

  7. [15]

    Animate anyone: Consistent and controllable image-to-video synthesis for character animation

    Li Hu, Xin Gao, Peng Zhang, Ke Sun, Bang Zhang, and Liefeng Bo. Animate anyone: Consistent and controllable image-to-video synthesis for character animation. arXiv preprint arXiv:2311.17117, 2023. 3

    2023 arXiv

  8. [16]

    Reconet: Recurrent correc- tion network for fast and efficient multi-modality image fu- sion

    Zhanbo Huang, Jinyuan Liu, Xin Fan, Risheng Liu, Wei Zhong, and Zhongxuan Luo. Reconet: Recurrent correc- tion network for fast and efficient multi-modality image fu- sion. In European Conference on Computer Vision , pages 539–555, 2022. 1

    2022

  9. [17]

    Multispectral pedestrian detection: Benchmark dataset and baseline

    Soonmin Hwang, Jaesik Park, Namil Kim, Yukyung Choi, and In So Kweon. Multispectral pedestrian detection: Benchmark dataset and baseline. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 1037–1045, 2015. 2, 5, 8, 9

    2015

  10. [18]

    A comprehensive review on in- telligent surveillance systems

    Sutrisno Warsono Ibrahim. A comprehensive review on in- telligent surveillance systems. Communications in science and technology, 1(1), 2016. 1, 8

    2016

  11. [19]

    Low-light image enhancement with wavelet-based diffusion models

    Hai Jiang, Ao Luo, Haoqiang Fan, Songchen Han, and Shuaicheng Liu. Low-light image enhancement with wavelet-based diffusion models. ACM Transactions on Graphics, 42(6):1–14, 2023. 3

    2023

  12. [20]

    Enlightengan: Deep light enhancement without paired supervision

    Yifan Jiang, Xinyu Gong, Ding Liu, Yu Cheng, Chen Fang, Xiaohui Shen, Jianchao Yang, Pan Zhou, and Zhangyang Wang. Enlightengan: Deep light enhancement without paired supervision. IEEE Transactions on Image Process- ing, 30:2340–2349, 2021. 1, 2, 5, 6, 7, 8, 9, 10, 11

    2021

  13. [21]

    Unsupervised night image enhancement: When layer decomposition meets light-effects suppression

    Yeying Jin, Wenhan Yang, and Robby T Tan. Unsupervised night image enhancement: When layer decomposition meets light-effects suppression. In European Conference on Com- puter Vision, pages 404–421, 2022. 1, 2

    2022

  14. [22]

    Iterative multi-granular image editing using diffusion models

    KJ Joseph, Prateksha Udhayanan, Tripti Shukla, Aishwarya Agarwal, Srikrishna Karanam, Koustava Goswami, and Bal- aji Vasan Srinivasan. Iterative multi-granular image editing using diffusion models. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vi...

    2024

  15. [23]

    A deep learning based image enhancement ap- proach for autonomous driving at night

    Guofa Li, Yifan Yang, Xingda Qu, Dongpu Cao, and Ke- qiang Li. A deep learning based image enhancement ap- proach for autonomous driving at night. Knowledge-Based Systems, 213:106617, 2021. 1

    2021

  16. [24]

    Different input resolutions and arbitrary output resolu- tion: A meta learning-based deep framework for infrared and visible image fusion

    Huafeng Li, Yueliang Cen, Yu Liu, Xun Chen, and Zhengtao Yu. Different input resolutions and arbitrary output resolu- tion: A meta learning-based deep framework for infrared and visible image fusion. IEEE Transactions on Image Process- ing, 30:4070–4083, 2021. 2

    2021

  17. [25]

    Densefuse: A fusion approach to infrared and visible images

    Hui Li and Xiao-Jun Wu. Densefuse: A fusion approach to infrared and visible images. IEEE Transactions on Image Processing, 28(5):2614–2623, 2018. 2

    2018

  18. [26]

    Crossfuse: A novel cross atten- tion mechanism based infrared and visible image fusion ap- proach

    Hui Li and Xiao-Jun Wu. Crossfuse: A novel cross atten- tion mechanism based infrared and visible image fusion ap- proach. Information Fusion, 103:102147, 2024. 1

    2024

  19. [27]

    Rfn-nest: An end-to- end residual fusion network for infrared and visible images

    Hui Li, Xiao-Jun Wu, and Josef Kittler. Rfn-nest: An end-to- end residual fusion network for infrared and visible images. Information Fusion, 73:72–86, 2021. 2

    2021

  20. [28]

    Controlnet++: Improving conditional controls with efficient consistency feedback

    Ming Li, Taojiannan Yang, Huafeng Kuang, Jie Wu, Zhaon- ing Wang, Xuefeng Xiao, and Chen Chen. Controlnet++: Improving conditional controls with efficient consistency feedback. arXiv preprint arXiv:2404.07987, 2024. 3

    2024 arXiv

  21. [29]

    Fusion from decomposition: A self-supervised decomposi- tion approach for image fusion

    Pengwei Liang, Junjun Jiang, Xianming Liu, and Jiayi Ma. Fusion from decomposition: A self-supervised decomposi- tion approach for image fusion. In European Conference on Computer Vision, pages 719–735, 2022. 1, 3, 5, 6, 7, 8, 9, 10

    2022

  22. [30]

    Mugan: thermal infrared image col- orization using mixed-skipping unet and generative adver- sarial network

    Hangying Liao, Qian Jiang, Xin Jin, Ling Liu, Lin Liu, Shin- Jye Lee, and Wei Zhou. Mugan: thermal infrared image col- orization using mixed-skipping unet and generative adver- sarial network. IEEE Transactions on Intelligent Vehicles , 8(4):2954–2969, 2023. 5, 6, 7, 8, 9, 10, 11

    2023

  23. [31]

    Image enhancement guided object detection in visually de- graded scenes

    Hongmin Liu, Fan Jin, Hui Zeng, Huayan Pu, and Bin Fan. Image enhancement guided object detection in visually de- graded scenes. IEEE transactions on neural networks and learning systems, 2023. 1

    2023

  24. [32]

    Rxdnfuse: A aggregated residual dense network for infrared and visible image fusion

    Yongzhi Long, Haitao Jia, Yida Zhong, Yadong Jiang, and Yuming Jia. Rxdnfuse: A aggregated residual dense network for infrared and visible image fusion. Information Fusion, 69:128–141, 2021. 2

    2021

  25. [33]

    Dpm-solver: A fast ode solver for diffusion probabilistic model sampling in around 10 steps

    Cheng Lu, Yuhao Zhou, Fan Bao, Jianfei Chen, Chongx- uan Li, and Jun Zhu. Dpm-solver: A fast ode solver for diffusion probabilistic model sampling in around 10 steps. Advances in Neural Information Processing Systems , pages 5775–5787, 2022. 3

    2022

  26. [34]

    Ddcgan: A dual-discriminator conditional gen- erative adversarial network for multi-resolution image fu- sion

    Jiayi Ma, Han Xu, Junjun Jiang, Xiaoguang Mei, and Xiao- Ping Zhang. Ddcgan: A dual-discriminator conditional gen- erative adversarial network for multi-resolution image fu- sion. IEEE Transactions on Image Processing , 29:4980– 4995, 2020. 2

    2020

  27. [35]

    Fusiongan: A generative adversarial net- work for infrared and visible image fusion

    Jiayi Ma, Wei Yu, Pengwei Liang, Chang Li, and Jun- jun Jiang Jiang. Fusiongan: A generative adversarial net- work for infrared and visible image fusion. Information Fu- sion, 48:11–26, 2019. 2

    2019

  28. [36]

    No-reference image quality assessment in the spatial domain

    Anish Mittal, Anush Krishna Moorthy, and Alan Con- rad Bovik. No-reference image quality assessment in the spatial domain. IEEE Transactions on image processing , 21(12):4695–4708, 2012. 7

    2012

  29. [37]

    completely blind

    Anish Mittal, Rajiv Soundararajan, and Alan C Bovik. Mak- ing a “completely blind” image quality analyzer. IEEE Sig- nal processing letters, 20(3):209–212, 2012. 6

    2012

  30. [38]

    Psenet: Progressive self-enhancement network for unsuper- vised extreme-light image enhancement

    Hue Nguyen, Diep Tran, Khoi Nguyen, and Rang Nguyen. Psenet: Progressive self-enhancement network for unsuper- vised extreme-light image enhancement. In Proceedings of the IEEE/CVF Winter Conference on Applications of Com- puter Vision, pages 1756–1765, 2023. 5, 6, 7, 8

    2023

  31. [39]

    High-resolution image syn- thesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj¨orn Ommer. High-resolution image syn- thesis with latent diffusion models. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 10684–10695, 2022. 3

    2022

  32. [40]

    Palette: Image-to-image diffusion models

    Chitwan Saharia, William Chan, Huiwen Chang, Chris Lee, Jonathan Ho, Tim Salimans, David Fleet, and Mohammad Norouzi. Palette: Image-to-image diffusion models. In ACM SIGGRAPH 2022 Conference Proceedings , pages 1– 10, 2022. 3

    2022

  33. [41]

    Progressive distillation for fast sampling of diffusion models

    Tim Salimans and Jonathan Ho. Progressive distillation for fast sampling of diffusion models. In International Confer- ence on Learning Representations, 2021. 3

    2021

  34. [42]

    Nighttime visibility en- hancement by increasing the dynamic range and suppres- sion of light effects

    Aashish Sharma and Robby T Tan. Nighttime visibility en- hancement by increasing the dynamic range and suppres- sion of light effects. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 11977– 11986, 2021. 1, 2

    2021

  35. [43]

    Difftalk: Crafting diffusion models for generalized audio-driven portraits animation

    Shuai Shen, Wenliang Zhao, Zibin Meng, Wanhua Li, Zheng Zhu, Jie Zhou, and Jiwen Lu. Difftalk: Crafting diffusion models for generalized audio-driven portraits animation. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 1982–1991, 2023. 2

    1982

  36. [44]

    Deep unsupervised learning using nonequilibrium thermodynamics

    Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. In International Confer- ence on Machine Learning, pages 2256–2265, 2015. 2, 3

    2015

  37. [45]

    Denois- ing diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denois- ing diffusion implicit models. In International Conference on Learning Representations, 2020. 3

    2020

  38. [46]

    Divfusion: Darkness-free infrared and visible im- age fusion

    Linfeng Tang, Xinyu Xiang, Hao Zhang, Meiqi Gong, and Jiayi Ma. Divfusion: Darkness-free infrared and visible im- age fusion. Information Fusion, 91:477–493, 2023. 2

    2023

  39. [47]

    Piafusion: A progressive infrared and visible im- age fusion network based on illumination aware.Information Fusion, 83:79–92, 2022

    Linfeng Tang, Jiteng Yuan, Hao Zhang, Xingyu Jiang, and Jiayi Ma. Piafusion: A progressive infrared and visible im- age fusion network based on illumination aware.Information Fusion, 83:79–92, 2022. 2, 5, 7, 9

    2022

  40. [48]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in Neural Information Processing Systems, 30, 2017. 3

    2017

  41. [49]

    Cs2fusion: Contrastive learning for self-supervised infrared and visible image fusion by esti- mating feature compensation map

    Xue Wang, Zheng Guan, Wenhua Qian, Jinde Cao, Shu Liang, and Jin Yan. Cs2fusion: Contrastive learning for self-supervised infrared and visible image fusion by esti- mating feature compensation map. Information Fusion , 102:102039, 2024. 1, 3

    2024

  42. [50]

    Uformer: A general u-shaped transformer for image restoration

    Zhendong Wang, Xiaodong Cun, Jianmin Bao, Wengang Zhou, Jianzhuang Liu, and Houqiang Li. Uformer: A general u-shaped transformer for image restoration. In IEEE Con- ference on Computer Vision and Pattern Recognition, pages 17662–17672, 2022. 10, 11

    2022

  43. [51]

    Classification saliency- based rule for visible and infrared image fusion.IEEE Trans- actions on Computational Imaging, 7:824–836, 2021

    Han Xu, Hao Zhang, and Jiayi Ma. Classification saliency- based rule for visible and infrared image fusion.IEEE Trans- actions on Computational Imaging, 7:824–836, 2021. 2

    2021

  44. [52]

    Learning enriched features for real image restoration and enhancement

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan, Ming-Hsuan Yang, and Ling Shao. Learning enriched features for real image restoration and enhancement. In European Conference on Computer Vi- sion, pages 492–511, 2020. 1, 2

    2020

  45. [53]

    Sdnet: A versatile squeeze-and- decomposition network for real-time image fusion

    Hao Zhang and Jiayi Ma. Sdnet: A versatile squeeze-and- decomposition network for real-time image fusion. Inter- national Journal of Computer Vision , 129(10):2761–2785,

  46. [54]

    Rethinking the image fusion: A fast unified image fusion network based on proportional maintenance of gradient and intensity

    Hao Zhang, Han Xu, Yang Xiao, Xiaojie Guo, and Jiayi Ma. Rethinking the image fusion: A fast unified image fusion network based on proportional maintenance of gradient and intensity. In Proceedings of the AAAI Conference on Artifi- cial Intelligence, pages 12797–12804, 2020. 2

    2020

  47. [55]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3836–3847, 2023. 2, 3

    2023

  48. [56]

    Cddfuse: Correlation-driven dual-branch feature decompo- sition for multi-modality image fusion

    Zixiang Zhao, Haowen Bai, Jiangshe Zhang, Yulun Zhang, Shuang Xu, Zudi Lin, Radu Timofte, and Luc Van Gool. Cddfuse: Correlation-driven dual-branch feature decompo- sition for multi-modality image fusion. InProceedings of the IEEE Conference on Computer Vision and Pattern Reco...

    2023

  49. [57]

    Equivariant multi-modality image fusion

    Zixiang Zhao, Haowen Bai, Jiangshe Zhang, Yulun Zhang, Kai Zhang, Shuang Xu, Dongdong Chen, Radu Timofte, and Luc Van Gool. Equivariant multi-modality image fusion. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 25912–25921, 2024....

    2024

  50. [58]

    Ddfm: Denoising diffusion model for multi-modality image fusion

    Zixiang Zhao, Haowen Bai, Yuanzhi Zhu, Jiangshe Zhang, Shuang Xu, Yulun Zhang, Kai Zhang, Deyu Meng, Radu Timofte, and Luc Van Gool. Ddfm: Denoising diffusion model for multi-modality image fusion. In Proceedings of the IEEE International Conference on Computer Vision , pages ...

    2023

  51. [59]

    Pyramid diffusion models for low-light image enhancement

    Dewei Zhou, Zongxin Yang, and Yi Yang. Pyramid diffusion models for low-light image enhancement. In Proceedings of the International Joint Conference on Artificial Intelligence, pages 1795–1803, 2023. 3, 5, 6, 7, 8, 9, 10, 11

    2023

  52. [60]

    Led- net: Joint low-light enhancement and deblurring in the dark

    Shangchen Zhou, Chongyi Li, and Chen Change Loy. Led- net: Joint low-light enhancement and deblurring in the dark. In European Conference on Computer Vision , pages 573– 589, 2022. 1

    2022

  53. [61]

    Deformable DETR: Deformable Transform- ers for End-to-End Object Detection

    Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, and Jifeng Dai. Deformable DETR: Deformable Transform- ers for End-to-End Object Detection. In International Con- ference on Learning Representations, 2020. 4

    2020

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.