Pith. sign in

REVIEW 4 major objections 6 minor 65 references

Natural Language Supervision for Low-light Image Enhancement

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read NaLSuper uses frozen CLIP text features, fed through cross-attention and attention fusion, to enhance low-light images; the paper reports state-of-the-art PSNR 24.01, SSIM 0.863, and LPIPS 0.0747 on LOLv1.

desk verdict First text-conditioned low-light enhancement paper, but the language signal is unspecified and its role is unproven without a control experiment. read the letter →

arxiv 2501.06546 v1 pith:FDGM3FD4 submitted 2025-01-11 cs.CV cs.AI

Jiahui Tang , Kaihua Zhou , Zhijian Luo , Yueen Hou This is my paper · ORCID
classification cs.CVcs.AI
keywords low-lightimageenhancementnaturallanguagesupervisioncross-modalattentionCLIPtextencodertext-guidedrestorationfusionLOLv1benchmarkvision-languagemodels
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 sets out to establish that natural language supervision—text descriptions of illumination injected into a low-light image enhancement network—can outperform conventional image-to-image supervision. It introduces NaLSuper, which combines a frozen CLIP text encoder with a Textual Guidance Conditioning Mechanism (TCM) that uses cross-attention to connect image regions to sentence words, plus an Information Fusion Attention (IFA) module that merges hierarchical image and text features. On the LOLv1 benchmark the paper reports PSNR 24.01, SSIM 0.863, and LPIPS 0.0747, ahead of all compared methods, and the best PSNR/SSIM on LOLv2-synthetic. If correct, this gives LLIE a new interface: enhancement can be steered by flexible text rather than by chasing a single 'perfect' reference image.

What carries the argument

The load-bearing object is the Textual Guidance Conditioning Mechanism (TCM), a cross-attention layer whose query comes from image features and whose key and value come from frozen CLIP text embeddings (Eqs. 1-2), designed to capture fine-grained alignment between image regions and sentence words. The second component is Information Fusion Attention (IFA), which combines channel attention, pixel attention, and a Cross-layer Attention Fusion Block (CAFB) that learns a 3-by-3 correlation matrix across hierarchical layers. Three Residual Textual guide Fusion Blocks stack TCM and IFA with residual connections, and the network is trained with L1 plus SSIM losses. The text side is a manually designed set of prompts (for example, 'normal light image') encoded by the fixed CLIP text encoder.

What would settle it

Train NaLSuper with a single fixed prompt for every image, or replace the text branch with a learned constant vector; if PSNR on LOLv1 remains near 24.01 dB, the reported gains do not depend on language content. Alternatively, measure the pairwise cosine similarity of the CLIP embeddings of all designed prompts: near-identical embeddings would show the text signal carries no per-image information.

Watch

Extended reading notes

Core claim

The central claim is that text features from a fixed CLIP encoder can act as a supervisory and conditioning signal for low-light image enhancement. NaLSuper passes a low-light image through three Residual Textual guide Fusion Blocks, each stacking TCM and IFA. TCM computes cross-attention in which queries come from image features and keys and values come from text features, letting words attend to image regions; IFA applies channel attention, pixel attention, and a Cross-layer Attention Fusion Block to merge multi-level features. Trained with L1 plus SSIM losses, the full model reaches 24.01 dB PSNR, 0.863 SSIM, and 0.0747 LPIPS on LOLv1, and 24.48 dB PSNR with 0.929 SSIM on LOLv2-synthetic; the ablation reports that both TCM and IFA contribute to these results.

Load-bearing premise

The load-bearing premise is that the hand-written prompts, passed through the frozen CLIP text encoder, carry image-specific guidance; because the paper specifies no per-image captions or prompt distribution, the text feature may be identical for every image, in which case TCM's cross-attention degenerates into a learned global bias.

Editorial extensions

If this is right

  • If the reported numbers hold, text-conditioned cross-modal attention becomes the state of the art on LOLv1 and LOLv2-synthetic, ahead of transformer-based restorers such as Restormer and LLFormer.
  • The ablation indicates that both TCM and IFA contribute: removing either module lowers PSNR on LOLv1 from the full model's 24.01 dB, so the gain is attributed to text conditioning plus multi-level fusion rather than to the backbone alone.
  • Because the text encoder is frozen and the prompts are generic, the method needs no per-image captions at inference, which makes the language-conditioned architecture simple to deploy on existing paired datasets.
  • The LPIPS drop (0.0747 versus 0.0905 for the next best listed method) suggests the text-guided fusion improves perceptual quality, not only pixel-level metrics.

Reading between the lines

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

  • A testable extension is to replace the fixed prompt set with per-image captions or varying illumination descriptions; if NaLSuper's gains grow, the language signal is genuinely image-specific rather than a constant conditioning vector.
  • If the CLIP embeddings of the designed prompts are nearly identical, then 'natural language supervision' as implemented reduces to a learned global bias; computing the pairwise similarity of $T_i$ across prompts would settle whether the text modality carries information.
  • The same TCM/IFA architecture could transfer to other image restoration tasks such as dehazing, deraining, or super-resolution by swapping the text prompts for task descriptions, since neither module is specific to low-light enhancement.
  • A prompt-paraphrase study, training on many phrasings and testing on held-out wording, would reveal how sensitive the enhancement is to language choice, a dimension the current paper does not explore.
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 / 6 minor

Summary. The paper proposes NaLSuper, a low-light image enhancement network that combines a frozen CLIP text encoder with a Textual Guidance Conditioning Mechanism (TCM) based on cross-attention and an Information Fusion Attention (IFA) module with channel, pixel, and cross-layer attention. The network is trained with an L1 plus SSIM loss and evaluated on LOLv1, LOLv2-real, LOLv2-synthetic, and SID. The authors report state-of-the-art results on LOLv1 (PSNR 24.01 dB, SSIM 0.863, LPIPS 0.0747) and LOLv2-synthetic, and attribute the gains to natural language supervision. The paper also includes ablations of the TCM and IFA components, hyperparameter experiments on the number of RTFB blocks, and a loss-function comparison.

Significance. If the reported results are reproducible, NaLSuper would be a new state of the art on LOLv1 and LOLv2-synthetic, and the TCM/IFA combination would be a useful architectural contribution. The paper's strengths are the breadth of the comparison on LOLv1 (20 methods, four metrics) and the component-level ablation. However, the central conceptual claim—that natural language supervision is the mechanism behind the gains—is not currently supported because no prompt set or text-conditioning control is reported. The numerical claims are also weakened by test-set-selected hyperparameters, missing error bars and code, unfilled placeholder values in the text, and an internal contradiction with Table II on SID and LOLv2-real.

major comments (4)
  1. [Section III-A.1, Eqs. (1)-(2)] The text prompt set and the image-to-prompt assignment rule are never specified; the only example given is "normal light image". If one prompt, or a small fixed set of prompts, is used for all images, then τθ(Ti) is constant across the batch, K and V in Eq. (1) are constant, and the cross-attention in Eq. (2) reduces to a learned global bias that can be absorbed by B and the projection matrices. In that case the "+TCM" gains in Table III show only the benefit of adding attention parameters, not of language semantics. The authors must list the full prompt templates and the per-image captioning protocol, and they must include a control experiment that replaces the text features with a constant token or a fixed random vector and reports PSNR/SSIM/LPIPS/MAE. Without this control, the central novelty claim of natural language supervision is not established.
  2. [Section IV-E.2, Tables III-IV] The hyperparameters (number of RTFB blocks and the loss combination) were selected by comparing on the LOLv1 test set: the text says "After comparison in the LOLv1 dataset, the number of 15 achieve the best performance", and the same test set is then used for the final reported scores. This is test-set selection, which can inflate the reported margin over competitors. In addition, Table III is introduced as reporting "average and standard deviation" but contains no standard deviations, and no code is released. The authors should select hyperparameters on a validation split, report variability over multiple runs or seeds, and release code and the prompt set so that the state-of-the-art claim can be independently verified.
  3. [Section IV-C, Table II] The text contains unfilled placeholders "—dB, —dB and —dB" for the gains over the second-best results on LOLv2-real, LOLv2-synthetic, and SID, and another "—" for the SSIM improvements. The same paragraph also claims the method achieves the highest PSNR on SID and on LOLv2-real, but Table II contradicts this: on SID, Restormer reports 22.27 dB versus NaLSuper's 22.22 dB, and on LOLv2-real, LLFormer reports 21.73 dB versus NaLSuper's 21.12 dB. This section must be completed and the quantitative claims reconciled with Table II.
  4. [Table III] The ablation does not isolate the language mechanism. The +TCM configuration is not uniformly beneficial: on LOLv2-real it reduces PSNR from 19.03 (Base) to 18.71, while the full model reaches 21.12. The large jump from +IFA (20.45) or +TCM (18.71) to +IFA+TCM (21.12) indicates a strong interaction, but the paper does not analyze this interaction or show that the text-conditioned branch, rather than the extra parameters, produces it. A matched-capacity image-only attention baseline and a constant-text control are needed to support the claim that TCM's textual guidance is the cause of the gains.
minor comments (6)
  1. [Eq. (5)] The output of the pixel-attention branch is written as CA(Fc); this should be PA(Fcout) or another symbol to distinguish it from the channel-attention output in Eq. (4).
  2. [Table I] The proposed method is labeled "NalSper" in the last row; it should read "NaLSuper".
  3. [Throughout] There are several typographical errors, including "wihch" in Section IV-C, "evolution" in Section IV-A, "Duadro" in Section IV-B, "a Information" in the Abstract, and "extensive experiment demonstrate" in the Abstract.
  4. [Section III-C] The notation is imprecise: Eq. (8) uses an unsubscripted norm, and the text says "µ is the variances, and σ is the covariance" while µ and σ are usually the mean and the standard deviation/covariance; please clarify.
  5. [Figures 5-7] The captions of Figures 5, 6, and 7 are identical; please indicate which dataset and which methods each figure shows.
  6. [Section IV-A, Table II] The paper states that four metrics are adopted, but Table II reports only PSNR and SSIM on LOLv2-real, LOLv2-synthetic, and SID; please either add LPIPS and MAE to Table II or state explicitly why they are omitted.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found; the text-conditioning degeneracy is a validity concern, not a circularity.

full rationale

NaLSuper does not claim a first-principles derivation; its claims are architectural and empirical. TCM is implemented as the standard cross-attention of Eq. (1)-(2), credited to Perceiver and Transformer, IFA is channel/pixel attention modeled on FFA-Net, and CAFB is adopted from LLFormer [46], none of which are self-citations of the present authors. No parameter is fitted and then renamed a prediction: the PSNR/SSIM/LPIPS/MAE scores are external measurements on benchmark test sets, and the ablation table is an ordinary component analysis. Hyperparameter selection on the test set is a generalization concern, not circularity. The one load-bearing conceptual premise is that the manually designed prompts ("such as normal light image") provide language guidance. If the same prompt is used for every image, Ti is constant and Eq. (1)-(2) reduce K and V to constants, so the +TCM gain could come from extra attention parameters rather than from linguistic content. That is a substantive validity gap in the interpretation, but it is not a circular derivation of the reported results from the prompts themselves, because the numbers are not constructed to equal the input by definition. Accordingly, under the circularity rubric, the honest finding is no significant circularity.

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

No invented physical entities are introduced; TCM and IFA are architectural components, not new forces, particles, or dimensions. The main unaccounted costs are the hand-chosen prompt set, test-set-selected hyperparameters, and domain assumptions that CLIP text features and cross-modal attention transfer from image-text matching to low-light restoration.

free parameters (5)
  • Number of RTFB blocks = 15
    Selected by ablating 3, 5, 8, 12, 15, and 20 blocks on the LOLv1 test set (Table IV); the final model uses 15.
  • Loss function combination = L1 + SSIM
    Chosen from L1, SSIM, and L1+SSIM using test-set results (Table V); exact weighting of the two terms is not reported.
  • Text prompt templates = not specified
    Prompts are manually designed (Section III-A.1), but the full prompt set and whether prompts vary per image are not given.
  • CAFB scaling factor delta
    Equation (6) includes a scaling factor delta, but its value or schedule is not reported.
  • Feature channel dimension C and attention head counts
    The main architecture description (Section III) leaves C and detailed attention dimensions unspecified.
assumptions (4)
  • domain assumption Frozen CLIP text embeddings of hand-designed prompts provide useful illumination-related guidance for low-light enhancement.
    Invoked in Section III-A.1; no evidence is given that the selected prompts correlate with scene content or illumination, and no per-image captions are described.
  • ad hoc to paper Cross-attention between image-region features and sentence-word features transfers from image-text matching to image restoration.
    Motivated by [51], an image-sentence matching method (Section III-A.2); the paper does not show that region-word alignment helps reconstruction.
  • domain assumption Ground-truth normal-light images are appropriate supervision targets despite the paper's own critique that no perfect reference exists.
    The introduction questions reference images (Section I), yet training and evaluation use L1 and SSIM against those references (Eqs. 8-10).
  • domain assumption Reported baseline numbers from re-run public code are accurate and directly comparable.
    Section IV-C says competitors were fine-tuned with recommended parameters, but no logs, seeds, or code are provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Natural Language Supervision for Low-light Image Enhancement." pith.science (2026). https://pith.science/paper/FDGM3FD4

@misc{pith2026250106546,
  author       = {Pith},
  title        = {Pith review of: Natural Language Supervision for Low-light Image Enhancement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FDGM3FD4}},
  note         = {Machine review of arXiv:2501.06546}
}
read the original abstract

With the development of deep learning, numerous methods for low-light image enhancement (LLIE) have demonstrated remarkable performance. Mainstream LLIE methods typically learn an end-to-end mapping based on pairs of low-light and normal-light images. However, normal-light images under varying illumination conditions serve as reference images, making it difficult to define a ``perfect'' reference image This leads to the challenge of reconciling metric-oriented and visual-friendly results. Recently, many cross-modal studies have found that side information from other related modalities can guide visual representation learning. Based on this, we introduce a Natural Language Supervision (NLS) strategy, which learns feature maps from text corresponding to images, offering a general and flexible interface for describing an image under different illumination. However, image distributions conditioned on textual descriptions are highly multimodal, which makes training difficult. To address this issue, we design a Textual Guidance Conditioning Mechanism (TCM) that incorporates the connections between image regions and sentence words, enhancing the ability to capture fine-grained cross-modal cues for images and text. This strategy not only utilizes a wider range of supervised sources, but also provides a new paradigm for LLIE based on visual and textual feature alignment. In order to effectively identify and merge features from various levels of image and textual information, we design an Information Fusion Attention (IFA) module to enhance different regions at different levels. We integrate the proposed TCM and IFA into a Natural Language Supervision network for LLIE, named NaLSuper. Finally, extensive experiments demonstrate the robustness and superior effectiveness of our proposed NaLSuper.

Figures

Figures reproduced from arXiv: 2501.06546 by the authors.

Figure 1
Figure 1. Comparison with state-of-the-art methods on LOLv1 dataset. It is [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview architecture of our proposed NaLSuper. NaLSuper is a Natural Language Supervision network for LLIE, which incorporates Textual [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Overview architecture of Information Fusion Attention (IFA). [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Overview architecture of Cross-layer Attention Fusion Block(CAFB). [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Visual comparison with LLIE methods on LOLv1 dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Visual comparison with LLIE methods on LOLv1 dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Visual comparison with LLIE methods on LOLv1 dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Visual comparison with LLIE methods on LOLv2-real dataset. [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Comparison of the subjective visual effects of models with varying configurations. [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Comparison of the subjective visual effects of models with different [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

65 extracted references · 52 canonical work pages

  1. [1]

    Low-light homomorphic filtering network for integrating image enhancement and classification

    Rayan Al Sobbahi and Joe Tekli. Low-light homomorphic filtering network for integrating image enhancement and classification. Signal Processing: Image Communication , 100:116527, 2022

    work page 2022

  2. [2]

    Video enhancement using per- pixel virtual exposures

    Eric P Bennett and Leonard McMillan. Video enhancement using per- pixel virtual exposures. In ACM SIGGRAPH 2005 Papers , pages 845–

    work page 2005

  3. [3]

    Learning to see in the dark

    Chen Chen, Qifeng Chen, Jia Xu, and Vladlen Koltun. Learning to see in the dark. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 3291–3300, 2018

    work page 2018

  4. [4]

    A simple and effective histogram equalization approach to image enhancement

    Heng-Da Cheng and XJ Shi. A simple and effective histogram equalization approach to image enhancement. Digital signal processing, 14(2):158–170, 2004

    work page 2004

  5. [5]

    Unifying vision- and-language tasks via text generation

    Jaemin Cho, Jie Lei, Hao Tan, and Mohit Bansal. Unifying vision- and-language tasks via text generation. In International Conference on Machine Learning, pages 1931–1942. PMLR, 2021. NATURAL LANGUAGE SUPERVISION FOR LOW-LIGHT IMAGE ENHANCEMENT, VOL. 18, NO. 9, AUGUST 2024 11

    work page 1931

  6. [6]

    Fast efficient algorithm for enhancement of low lighting video

    Xuan Dong, Yi Pang, and Jiangtao Wen. Fast efficient algorithm for enhancement of low lighting video. In ACM SIGGRApH 2010 posters , pages 1–1. 2010

    work page 2010

  7. [7]

    Clap learning audio concepts from natural language su- pervision

    Benjamin Elizalde, Soham Deshmukh, Mahmoud Al Ismail, and Huam- ing Wang. Clap learning audio concepts from natural language su- pervision. In ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages 1–5. IEEE, 2023

    work page 2023

  8. [8]

    A probabilistic method for image enhancement with simultaneous illumination and reflectance estimation

    Xueyang Fu, Yinghao Liao, Delu Zeng, Yue Huang, Xiao-Ping Zhang, and Xinghao Ding. A probabilistic method for image enhancement with simultaneous illumination and reflectance estimation. IEEE Transactions on Image Processing , 24(12):4965–4977, 2015

    work page 2015

Show all 65 references
  1. [9]

    A fusion-based enhancing method for weakly illuminated images

    Xueyang Fu, Delu Zeng, Yue Huang, Yinghao Liao, Xinghao Ding, and John Paisley. A fusion-based enhancing method for weakly illuminated images. Signal Processing, 129:82–96, 2016

    2016

  2. [10]

    A weighted variational model for simultaneous reflectance and illumination estimation

    Xueyang Fu, Delu Zeng, Yue Huang, Xiao-Ping Zhang, and Xinghao Ding. A weighted variational model for simultaneous reflectance and illumination estimation. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 2782–2790, 2016

    2016

  3. [11]

    Open-vocabulary object detection via vision and language knowledge distillation

    Xiuye Gu, Tsung-Yi Lin, Weicheng Kuo, and Yin Cui. Open-vocabulary object detection via vision and language knowledge distillation. arXiv preprint arXiv:2104.13921, 2021

    2021 arXiv

  4. [12]

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

    Chunle Guo, Chongyi Li, Jichang Guo, Chen Change Loy, Junhui Hou, Sam Kwong, and Runmin Cong. Zero-reference deep curve estimation for low-light image enhancement. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 1780– 1789, 2020

    2020

  5. [13]

    Lime: Low-light image enhance- ment via illumination map estimation

    Xiaojie Guo, Yu Li, and Haibin Ling. Lime: Low-light image enhance- ment via illumination map estimation. IEEE Transactions on image processing, 26(2):982–993, 2016

    2016

  6. [14]

    Low- light image enhancement with semi-decoupled decomposition

    Shijie Hao, Xu Han, Yanrong Guo, Xin Xu, and Meng Wang. Low- light image enhancement with semi-decoupled decomposition. IEEE transactions on multimedia , 22(12):3025–3038, 2020

    2020

  7. [15]

    Single image haze removal using dark channel prior

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

    2010

  8. [16]

    Semantic segmentation of underwater imagery: Dataset and benchmark

    Md Jahidul Islam, Chelsey Edge, Yuyang Xiao, Peigen Luo, Muntaqim Mehtaz, Christopher Morse, Sadman Sakib Enan, and Junaed Sattar. Semantic segmentation of underwater imagery: Dataset and benchmark. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (I...

    2020

  9. [17]

    Perceiver: General perception with iterative attention

    Andrew Jaegle, Felix Gimeno, Andy Brock, Oriol Vinyals, Andrew Zisserman, and Joao Carreira. Perceiver: General perception with iterative attention. In International conference on machine learning , pages 4651–4664. PMLR, 2021

    2021

  10. [18]

    Scaling up visual and vision-language representation learning with noisy text supervision

    Chao Jia, Yinfei Yang, Ye Xia, Yi-Ting Chen, Zarana Parekh, Hieu Pham, Quoc Le, Yun-Hsuan Sung, Zhen Li, and Tom Duerig. Scaling up visual and vision-language representation learning with noisy text supervision. In International conference on machine learning , pages 4904–4916...

    2021

  11. [19]

    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 processing , 30:2340–2349, 2021

    2021

  12. [20]

    A good prompt is worth millions of parameters: Low-resource prompt-based learning for vision-language models

    Woojeong Jin, Yu Cheng, Yelong Shen, Weizhu Chen, and Xiang Ren. A good prompt is worth millions of parameters: Low-resource prompt-based learning for vision-language models. arXiv preprint arXiv:2110.08484, 2021

    2021 arXiv

  13. [21]

    A multiscale retinex for bridging the gap between color images and the human obser- vation of scenes

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

    1997

  14. [22]

    Prompting visual-language models for efficient video understanding

    Chen Ju, Tengda Han, Kunhao Zheng, Ya Zhang, and Weidi Xie. Prompting visual-language models for efficient video understanding. In European Conference on Computer Vision , pages 105–124. Springer, 2022

    2022

  15. [23]

    Contrast enhancement using brightness preserving bi-histogram equalization

    Yeong-Taeg Kim. Contrast enhancement using brightness preserving bi-histogram equalization. IEEE transactions on Consumer Electronics , 43(1):1–8, 1997

    1997

  16. [24]

    Madnet: a fast and lightweight network for single-image super resolution

    Rushi Lan, Long Sun, Zhenbing Liu, Huimin Lu, Cheng Pang, and Xiaonan Luo. Madnet: a fast and lightweight network for single-image super resolution. IEEE transactions on cybernetics , 51(3):1443–1453, 2020

    2020

  17. [25]

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

    Chongyi Li, Chunle Guo, and Chen Change Loy. Learning to enhance low-light image via zero-reference deep curve estimation. IEEE Trans- actions on Pattern Analysis and Machine Intelligence, 44(8):4225–4238, 2021

    2021

  18. [26]

    Lightennet: A convolutional neural network for weakly illuminated image enhance- ment

    Chongyi Li, Jichang Guo, Fatih Porikli, and Yanwei Pang. Lightennet: A convolutional neural network for weakly illuminated image enhance- ment. Pattern recognition letters, 104:15–22, 2018

    2018

  19. [27]

    Low-light image enhance- ment via progressive-recursive network

    Jinjiang Li, Xiaomei Feng, and Zhen Hua. Low-light image enhance- ment via progressive-recursive network. IEEE Transactions on Circuits and Systems for Video Technology , 31(11):4227–4240, 2021

    2021

  20. [28]

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

    Seokjae Lim and Wonjun Kim. Dslr: Deep stacked laplacian restorer for low-light image enhancement. IEEE Transactions on Multimedia , 23:4272–4284, 2020

    2020

  21. [29]

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

    Risheng Liu, Long Ma, Jiaao Zhang, Xin Fan, and Zhongxuan Luo. Retinex-inspired unrolling with cooperative prior architecture search for low-light image enhancement. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10561– 10570, 2021

    2021

  22. [30]

    Ssd: Single shot multibox detector

    Wei Liu, Dragomir Anguelov, Dumitru Erhan, Christian Szegedy, Scott Reed, Cheng-Yang Fu, and Alexander C Berg. Ssd: Single shot multibox detector. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14, p...

    2016

  23. [31]

    Llnet: A deep autoencoder approach to natural low-light image enhancement

    Kin Gwn Lore, Adedotun Akintayo, and Soumik Sarkar. Llnet: A deep autoencoder approach to natural low-light image enhancement. Pattern Recognition, 61:650–662, 2017

    2017

  24. [32]

    Learning deep context-sensitive decomposition for low-light image enhancement

    Long Ma, Risheng Liu, Jiaao Zhang, Xin Fan, and Zhongxuan Luo. Learning deep context-sensitive decomposition for low-light image enhancement. IEEE Transactions on Neural Networks and Learning Systems, 33(10):5666–5680, 2021

    2021

  25. [33]

    Contrast-limited adaptive histogram equalization: Speed and effectiveness stephen m

    Stephen M Pizer. Contrast-limited adaptive histogram equalization: Speed and effectiveness stephen m. pizer, r. eugene johnston, james p. ericksen, bonnie c. yankaskas, keith e. muller medical image display research group. In Proceedings of the first conference on visualizatio...

    1990

  26. [34]

    Ffa-net: Feature fusion attention network for single image dehazing

    Xu Qin, Zhilin Wang, Yuanchao Bai, Xiaodong Xie, and Huizhu Jia. Ffa-net: Feature fusion attention network for single image dehazing. In Proceedings of the AAAI conference on artificial intelligence, volume 34, pages 11908–11915, 2020

    2020

  27. [35]

    Learning transferable visual models from natural language supervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervision. In International conference on machine learning , p...

    2021

  28. [36]

    Denseclip: Language-guided dense prediction with context-aware prompting

    Yongming Rao, Wenliang Zhao, Guangyi Chen, Yansong Tang, Zheng Zhu, Guan Huang, Jie Zhou, and Jiwen Lu. Denseclip: Language-guided dense prediction with context-aware prompting. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 18082–1...

    2022

  29. [37]

    Low-light image enhancement via a deep hybrid network

    Wenqi Ren, Sifei Liu, Lin Ma, Qianqian Xu, Xiangyu Xu, Xiaochun Cao, Junping Du, and Ming-Hsuan Yang. Low-light image enhancement via a deep hybrid network. IEEE Transactions on Image Processing , 28(9):4364–4375, 2019

    2019

  30. [38]

    How much can clip benefit vision-and-language tasks? arXiv preprint arXiv:2107.06383 , 2021

    Sheng Shen, Liunian Harold Li, Hao Tan, Mohit Bansal, Anna Rohrbach, Kai-Wei Chang, Zhewei Yao, and Kurt Keutzer. How much can clip benefit vision-and-language tasks? arXiv preprint arXiv:2107.06383 , 2021

    2021 arXiv

  31. [39]

    Ll-unet++: Unet++ based nested skip connections network for low-light image enhancement

    Pengfei Shi, Xiwang Xu, Xinnan Fan, Xudong Yang, and Yuanxue Xin. Ll-unet++: Unet++ based nested skip connections network for low-light image enhancement. IEEE Transactions on Computational Imaging , 2024

    2024

  32. [40]

    Automatically designing cnn architectures using the genetic algorithm for image classification

    Yanan Sun, Bing Xue, Mengjie Zhang, Gary G Yen, and Jiancheng Lv. Automatically designing cnn architectures using the genetic algorithm for image classification. IEEE transactions on cybernetics , 50(9):3840– 3854, 2020

    2020

  33. [41]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems , 30, 2017

    2017

  34. [42]

    Light- ening network for low-light image enhancement

    Li-Wen Wang, Zhi-Song Liu, Wan-Chi Siu, and Daniel PK Lun. Light- ening network for low-light image enhancement. IEEE Transactions on Image Processing, 29:7984–7996, 2020

    2020

  35. [43]

    Actionclip: A new paradigm for video action recognition

    Mengmeng Wang, Jiazheng Xing, and Yong Liu. Actionclip: A new paradigm for video action recognition. arXiv preprint arXiv:2109.08472, 2021

    2021 arXiv

  36. [44]

    Naturalness preserved enhancement algorithm for non-uniform illumination images

    Shuhang Wang, Jin Zheng, Hai-Miao Hu, and Bo Li. Naturalness preserved enhancement algorithm for non-uniform illumination images. IEEE transactions on image processing , 22(9):3538–3548, 2013

    2013

  37. [45]

    Learning transferable human-object interaction detector with natural language supervision

    Suchen Wang, Yueqi Duan, Henghui Ding, Yap-Peng Tan, Kim-Hui Yap, and Junsong Yuan. Learning transferable human-object interaction detector with natural language supervision. In Proceedings of the NATURAL LANGUAGE SUPERVISION FOR LOW-LIGHT IMAGE ENHANCEMENT, VOL. 18, NO. 9, AU...

    2024

  38. [46]

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

    Tao Wang, Kaihao Zhang, Tianrun Shen, Wenhan Luo, Bjorn Stenger, and Tong Lu. Ultra-high-definition low-light image enhancement: A benchmark and transformer-based method. In Proceedings of the AAAI Conference on Artificial Intelligence , volume 37, pages 2654–2662, 2023

    2023

  39. [47]

    Image enhancement based on equal area dualistic sub-image histogram equalization method

    Yu Wang, Qian Chen, and Baeomin Zhang. Image enhancement based on equal area dualistic sub-image histogram equalization method. IEEE transactions on Consumer Electronics , 45(1):68–75, 1999

    1999

  40. [48]

    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 Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 17683– 17693, 2022

    2022

  41. [49]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing , 13(4):600–612, 2004

    2004

  42. [50]

    Deep retinex decomposition for low-light enhancement

    Chen Wei, Wenjing Wang, Wenhan Yang, and Jiaying Liu. Deep retinex decomposition for low-light enhancement. arXiv preprint arXiv:1808.04560, 2018

    2018 arXiv

  43. [51]

    Multi- modality cross attention network for image and sentence matching

    Xi Wei, Tianzhu Zhang, Yan Li, Yongdong Zhang, and Feng Wu. Multi- modality cross attention network for image and sentence matching. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10941–10950, 2020

    2020

  44. [52]

    Uretinex-net: Retinex-based deep unfolding network for low-light image enhancement

    Wenhui Wu, Jian Weng, Pingping Zhang, Xu Wang, Wenhan Yang, and Jianmin Jiang. Uretinex-net: Retinex-based deep unfolding network for low-light image enhancement. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 5901– 5910, 2022

    2022

  45. [53]

    Learning to restore low-light images via decomposition-and-enhancement

    Ke Xu, Xin Yang, Baocai Yin, and Rynson WH Lau. Learning to restore low-light images via decomposition-and-enhancement. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2281–2290, 2020

    2020

  46. [54]

    Snr- aware low-light image enhancement

    Xiaogang Xu, Ruixing Wang, Chi-Wing Fu, and Jiaya Jia. Snr- aware low-light image enhancement. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17714– 17724, 2022

    2022

  47. [55]

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

    Wenhan Yang, Shiqi Wang, Yuming Fang, Yue Wang, and Jiaying Liu. From fidelity to perceptual quality: A semi-supervised approach for low- light image enhancement. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 3063–3072, 2020

    2020

  48. [56]

    A bio-inspired multi-exposure fusion framework for low-light image enhancement

    Zhenqiang Ying, Ge Li, and Wen Gao. A bio-inspired multi-exposure fusion framework for low-light image enhancement. arXiv preprint arXiv:1711.00591, 2017

    2017 arXiv

  49. [57]

    Restormer: Efficient transformer for high-resolution image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan, and Ming-Hsuan Yang. Restormer: Efficient transformer for high-resolution image restoration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 5728–5739, 2022

    2022

  50. [58]

    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 Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proce...

    2020

  51. [59]

    Be- yond brightening low-light images

    Yonghua Zhang, Xiaojie Guo, Jiayi Ma, Wei Liu, and Jiawan Zhang. Be- yond brightening low-light images. International Journal of Computer Vision, 129:1013–1037, 2021

    2021

  52. [60]

    Kindling the darkness: A practical low-light image enhancer

    Yonghua Zhang, Jiawan Zhang, and Xiaojie Guo. Kindling the darkness: A practical low-light image enhancer. In Proceedings of the 27th ACM international conference on multimedia , pages 1632–1640, 2019

    2019

  53. [61]

    Deep color consistent network for low-light image enhancement

    Zhao Zhang, Huan Zheng, Richang Hong, Mingliang Xu, Shuicheng Yan, and Meng Wang. Deep color consistent network for low-light image enhancement. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 1899–1908, 2022

    1908

  54. [62]

    Retinexdip: A unified deep framework for low-light image enhancement

    Zunjin Zhao, Bangshu Xiong, Lei Wang, Qiaofeng Ou, Lei Yu, and Fa Kuang. Retinexdip: A unified deep framework for low-light image enhancement. IEEE Transactions on Circuits and Systems for Video Technology, 32(3):1076–1088, 2021

    2021

  55. [63]

    Learning to generate scene graph from natural language supervision

    Yiwu Zhong, Jing Shi, Jianwei Yang, Chenliang Xu, and Yin Li. Learning to generate scene graph from natural language supervision. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 1823–1834, 2021

    2021

  56. [64]

    Learning to prompt for vision-language models

    Kaiyang Zhou, Jingkang Yang, Chen Change Loy, and Ziwei Liu. Learning to prompt for vision-language models. International Journal of Computer Vision , 130(9):2337–2348, 2022

    2022

  57. [65]

    Enlightening low-light images with dynamic guidance for context enrichment

    Lingyu Zhu, Wenhan Yang, Baoliang Chen, Fangbo Lu, and Shiqi Wang. Enlightening low-light images with dynamic guidance for context enrichment. IEEE Transactions on Circuits and Systems for Video Technology, 32(8):5068–5079, 2022

    2022

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.