Pith. sign in

REVIEW 4 major objections 7 minor 30 references

RT-X Net: RGB-Thermal cross attention network for Low-Light Image Enhancement

T0 review · 4 major / 7 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read RT-X Net fuses RGB and thermal images with cross-attention, reporting state-of-the-art low-light enhancement on LLVIP and the new V-TIEE benchmark.

desk verdict RT-X Net is a plausible, modest empirical contribution—new thermal cross-attention for low-light enhancement plus a fresh dataset—but its main LLVIP numbers rest on a small synthetic benchmark, so the real-world benefit needs firmer evidence. read the letter →

arxiv 2505.24705 v1 pith:ZSIZADWB submitted 2025-05-30 cs.CV

Raman Jha , Adithya Lenka , Mani Ramanagopal , Aswin Sankaranarayanan , Kaushik Mitra This is my paper · ORCID
classification cs.CV
keywords low-lightimageenhancementRGB-thermalfusioncross-attentiontransformerthermalimagingrestorationLLVIPdatasetV-TIEE
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

Low-light image enhancement usually works on RGB alone, but at night the visible image is dominated by noise and glare while a thermal camera still sees clean structure. The paper tries to establish that a transformer that fuses RGB and thermal images through cross-attention can beat RGB-only state-of-the-art methods, and that the fusion mechanism matters more than simply having the thermal data. If this claim is right, it gives surveillance, automotive, and robotics systems with thermal sensors a concrete way to produce sharper, better-exposed visible images at night, and it provides a new benchmark dataset, V-TIEE, for testing such methods on real nighttime scenes.

What carries the argument

The load-bearing component is the multi-head cross-attention module (MCA), which fuses the illumination-guided self-attention features of the RGB and thermal branches. In MCA, queries are taken from one modality while keys and values come from the other, so the thermal stream can reweight and refine the RGB feature map; the module is preceded by a Retinex-based illumination estimator that supplies an illumination map and illumination features used to reweight attention values. A PCA step reduces the fused channel dimensionality, and the enhanced image is produced by an MLP reconstruction network trained with mean absolute error.

What would settle it

Take RT-X Net trained on LLVIP and evaluate it on V-TIEE with the thermal image deliberately shifted by 1, 3, and 5 pixels relative to the visible image. If the PSNR or LPIPS advantage over the RGB-only ablation shrinks or reverses with small shifts, the method's real-world benefit depends on near-perfect alignment rather than on thermal structure per se.

Watch

Extended reading notes

Core claim

RT-X Net is a transformer for low-light image enhancement that fuses a visible RGB image with a co-registered thermal image. The paper's central claim is that thermal imagery, which records scene structure independently of illumination, supplies the texture and edge cues that RGB-only enhancers miss, and that a cross-attention module is the mechanism that lets thermal features guide the RGB representation. On the LLVIP dataset the paper reports a peak signal-to-noise ratio of 27.75 dB and SSIM of 0.85, against 26.59 dB and 0.79 for Retinexformer, and on the newly introduced V-TIEE real-world dataset it reports LPIPS of 0.12 and SSIM of 0.71, also the best among the compared methods. Ablation experiments show that removing the cross-attention and using RGB only drops PSNR by more than 1.3 dB, while simply concatenating the thermal channel yields a smaller gain.

Load-bearing premise

The real-world V-TIEE evaluation assumes that the visible and thermal cameras, which have different resolutions and fields of view, are aligned accurately enough at every pixel by homography for the cross-attention to fuse corresponding locations.

Editorial extensions

If this is right

  • Systems that already carry thermal cameras, such as autonomous vehicles, drones, and surveillance rigs, can use cross-attention fusion to produce visible-light images that are sharper and better exposed than RGB-only enhancers.
  • The V-TIEE dataset provides a real-world benchmark with genuine noise and exposure variation, so future methods can be tested without relying on simulated darkening of daytime images.
  • The fusion design transfers to other illumination-invariant modalities such as depth or event streams, where one sensor is robust to darkness and the other supplies color or dense context.
  • At 0.67 million parameters, the model is compact enough to make the architecture a candidate for embedded or mobile deployment.

Reading between the lines

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

  • The LLVIP evaluation darkens images by simulated exposure reduction and added noise; real nighttime glare and sensor noise may make the thermal advantage larger than the 1.16 dB PSNR gain reported here.
  • Because the gain over Retinexformer is modest, the practical value of RT-X Net depends on the platform already having a co-located thermal sensor; for RGB-only systems the method offers no benefit.
  • A direct experiment shifting or blurring the thermal input would show whether the improvement comes from high-frequency thermal edges or from coarse brightness guidance, and whether the cross-attention is robust to misalignment.
  • The V-TIEE benchmark could double as an alignment testbed: reporting PSNR and LPIPS under deliberately misregistered thermal inputs would quantify how much of the real-world gain relies on the homography refinement.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The paper proposes RT-X Net, a transformer-based network that fuses RGB and thermal images for low-light image enhancement. The architecture extracts illumination-guided self-attention features from both modalities and fuses them with a multi-head cross-attention module, followed by a reconstruction network. The authors introduce a real-world Visible-Thermal Image Enhancement Evaluation (V-TIEE) dataset with 50 co-located visible and thermal scenes, and they evaluate RT-X Net on a subset of the LLVIP dataset and on V-TIEE. Quantitative results in Table 1 report PSNR 27.75 dB / SSIM 0.85 on LLVIP, outperforming Retinexformer by 1.16 dB, and LPIPS 0.12 / SSIM 0.71 on V-TIEE. The ablation in Table 2 shows that cross-attention fusion outperforms both RGB-only self-attention and thermal channel concatenation. The central claim is that cross-modal attention to thermal structure yields a meaningful, though modest, gain for nighttime enhancement.

Significance. If the reported results are robust, the paper makes a useful contribution to low-light image enhancement by demonstrating that thermal imagery can be effectively integrated through cross-attention. The proposed architecture is lightweight (0.67M parameters) and the ablation supports the design choice of cross-attention over concatenation. The release of code and the V-TIEE dataset is also a concrete contribution to the community. However, the quantitative evidence for the central claim is weakened by the synthetic nature of the main LLVIP evaluation, the small size of the real-world V-TIEE set, the lack of error bars, and the absence of the most relevant thermal-guided baseline. These issues do not invalidate the approach, but they need to be addressed before the claim that RT-X Net outperforms state-of-the-art methods is fully convincing.

major comments (4)
  1. [§4.2.1, §4.2.2, Table 1] The LLVIP evaluation appears to measure how well each method inverts the same synthetic low-light degradation family used in training. Section 4.2.1 states that low-exposure conditions are simulated by reducing exposure (×5–20) and adding noise, and Section 4.2.2 states that the model is trained on the same simulated low-light pairs with an MAE loss. The reported PSNR/SSIM gains, including the 1.16 dB improvement over Retinexformer, could therefore reflect better fitting to this specific synthetic corruption model rather than a genuine benefit of thermal structure in real night scenes. Please provide evidence that the improvement persists under a different degradation protocol, on real low-light RGB-thermal pairs with ground truth, or at least with a held-out degradation setting not seen in training.
  2. [§4.1, §4.4, V-TIEE results] The real-world evidence on V-TIEE is too thin to carry the generalization claim. The dataset contains only 50 scenes, no error bars or per-scene statistics are reported, and the visible and thermal sensors have different resolutions (1440×1080 vs. 640×512) and fields of view (34° vs. 24° HFOV), with alignment refined by a homography. Residual parallax or homography error would mix content from non-corresponding spatial locations, which is especially problematic for a cross-attention mechanism that assumes corresponding positions across modalities. Please report per-scene or bootstrap error bars, describe the alignment accuracy quantitatively, and show that the observed gains are robust to alignment perturbation.
  3. [§2.2, §4.3, Table 1] The most relevant thermal-guided baseline, Cao et al. [20], is not compared because its code is unavailable. This baseline directly targets thermal-guided low-light visible image enhancement, so its absence means the comparison does not isolate the benefit of the proposed cross-attention fusion mechanism. Please include a re-implementation of [20] on the same protocol, or provide a clear quantitative and qualitative comparison that justifies why the omission does not affect the claimed superiority.
  4. [§4.5, Table 2] The ablation study supports the cross-attention module, but it is conducted solely on the synthetic LLVIP protocol. The 0.6 dB gain over thermal channel concatenation may not transfer to real low-light conditions, where alignment errors and domain shift are present. Please report the same ablation on a real-world evaluation set, even if only with perceptual metrics, to demonstrate that the design choice is beneficial in the intended application setting.
minor comments (7)
  1. [Title and header] The title line in the paper header contains a typo: 'CROSS A TTENTION' should read 'CROSS ATTENTION'.
  2. [§3.1, Eq. (2)] Equation (2) defines the illumination estimator as a function of a single input I, but the text says it predicts from both RGB and thermal images. Please clarify whether the illumination estimator is shared across modalities or whether two separate estimators are used.
  3. [§3.2, Eq. (3)] The notation SelfAttn(xl, Fillum) suggests that illumination features are an input to the self-attention operator, but the text only says they are used as a learnable re-weighting term on the values. Please specify the exact integration mechanism in the equation or the surrounding text.
  4. [§3.3, Eq. (5)] The cross-attention module is written as MCA(xa_RGB ∥ xa_Therm), but the text says 'queries are derived from one modality, while keys and values come from the other.' With a concatenated input it is unclear which projection produces queries and keys from which branch. Please define the query, key, and value projections precisely.
  5. [§3.3] The PCA step after cross-attention is mentioned but not described; please state how the PCA is computed, whether it is learned or fixed, and what dimensionality is used.
  6. [§4.2.1, Table 1] Table 1 reports a single value per metric without error bars. Given that the LLVIP test set has only 45 images and V-TIEE has 50 scenes, standard deviations or confidence intervals are needed to assess whether the reported differences are statistically meaningful.
  7. [§4.2.1] The noise addition is attributed to Hasinoff et al. [22], but the exact noise model and parameters are not specified. Please describe how the simulated low-light images are generated so the experiments are reproducible.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: RT-X Net is trained on LLVIP simulated low-light pairs and evaluated on held-out LLVIP and real-world V-TIEE data; no load-bearing step reduces to its inputs.

full rationale

The paper's central claim is an empirical architecture comparison, not a derivation from inputs to a predicted target. The RGB and thermal branches are trained on LLVIP pairs synthesized by exposure reduction and noise addition (Section 4.2.1: 'Low-exposure conditions are simulated by reducing exposure (×5–20) and adding noise [22]'), and the reported LLVIP numbers are measured on a held-out 45-image test subset using PSNR/SSIM; no metric is fitted or reused as a prediction. The V-TIEE evaluation is an independent real-world generalization check with high-exposure reference images, and the paper explicitly notes that the closest thermal-guided baseline (Cao et al. [20]) was unavailable for comparison, which weakens but does not make circular the comparison. The ablation (Table 2) tests cross-attention against RGB-only self-attention and thermal channel concatenation on the same protocol, so it is an internal ablation, not a self-referential construction. The paper contains no self-citations used as load-bearing justification, no imported uniqueness theorem, and no ansatz disguised as derivation. The main limitation—simulated LLVIP degradation may not fully represent real night scenes—is a correctness or generalization concern, not circularity, because the model is not predicting its own fitted targets. Verdict: no significant circularity.

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

The central claim rests on the empirical premise that thermal structure aligns with visible content and that simulated degradation is representative. No new physical entities are introduced. The main free parameters are learned network weights; the degradation range and hyperparameters are hand-chosen but standard.

free parameters (3)
  • Network weights of RT-X Net = 0.67 million parameters
    All model parameters are learned by minimizing MAE on LLVIP training pairs; this is the normal fitted content of a deep network.
  • Simulated low-exposure degradation factors = x5-20 exposure reduction plus added noise
    Inputs for LLVIP training and testing are produced by this hand-chosen simulation range, which may influence the relative ordering of methods.
  • Training hyperparameters = learning rate 2e-4, batch size 4, patch size 128, 1.5e5 iterations
    Manually selected standard choices; they are not fitted to the test metrics.
assumptions (4)
  • domain assumption Retinex decomposition I = R * M holds for low-light images, with reflectance and illumination cleanly separable.
    Introduced in Section 3.1; the illumination estimator predicts M and the network amplifies the lit image by M. If real low-light images do not factor cleanly, the enhancement premise weakens.
  • domain assumption Thermal images provide structurally informative cues that are spatially aligned with visible scene content.
    Used in Sections 3.2 and 3.3 to justify using thermal self-attended features as guidance for RGB reconstruction; it is an empirical premise, not proven.
  • domain assumption Simulated low-exposure inputs, exposure reduction by x5-20 plus noise, are representative of real nighttime degradations.
    The LLVIP evaluation in Section 4.2.1 relies on this to claim state-of-the-art performance; generalization to V-TIEE is the only real-world check.
  • domain assumption Paired visible-thermal images are co-located and homography-aligned well enough for cross-modal attention.
    V-TIEE dataset construction in Section 4.1 depends on this; alignment error would break cross-attention correspondences.

how reviews work

0 comments
Cite this review

Pith. "Pith review of RT-X Net: RGB-Thermal cross attention network for Low-Light Image Enhancement." pith.science (2026). https://pith.science/paper/ZSIZADWB

@misc{pith2026250524705,
  author       = {Pith},
  title        = {Pith review of: RT-X Net: RGB-Thermal cross attention network for Low-Light Image Enhancement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZSIZADWB}},
  note         = {Machine review of arXiv:2505.24705}
}
read the original abstract

In nighttime conditions, high noise levels and bright illumination sources degrade image quality, making low-light image enhancement challenging. Thermal images provide complementary information, offering richer textures and structural details. We propose RT-X Net, a cross-attention network that fuses RGB and thermal images for nighttime image enhancement. We leverage self-attention networks for feature extraction and a cross-attention mechanism for fusion to effectively integrate information from both modalities. To support research in this domain, we introduce the Visible-Thermal Image Enhancement Evaluation (V-TIEE) dataset, comprising 50 co-located visible and thermal images captured under diverse nighttime conditions. Extensive evaluations on the publicly available LLVIP dataset and our V-TIEE dataset demonstrate that RT-X Net outperforms state-of-the-art methods in low-light image enhancement. The code and the V-TIEE can be found here https://github.com/jhakrraman/rt-xnet.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 28 canonical work pages

  1. [20]

    Learning to see in the dark,

    Chen Chen, Qifeng Chen, Jia Xu, and Vladlen Koltun, “Learning to see in the dark,” in IEEE CVPR, 2018, pp. 3291–3300

    work page 2018

  2. [1]

    INTRODUCTION Image enhancement is crucial in various domains, such as medical imaging, surveillance, and environmental monitor- ing, as it improves image quality for more accurate informa- tion and better decision-making outcomes. Building on the success of convolutional neural networks (CNNs) in image denoising [1], CNN-based architectures have also been...

    arXiv 2025

  3. [2]

    Low-Light Image Enhancement Low-light image enhancement is challenging due to dynamic illumination and noise in nighttime conditions

    PREVIOUS WORK 2.1. Low-Light Image Enhancement Low-light image enhancement is challenging due to dynamic illumination and noise in nighttime conditions. Traditional methods like histogram stretching [12] had limited success. In contrast, CNN-based approaches [13] have improved results leveraging the retinex theory [5, 6] to use illumination maps and refle...

    work page

  4. [3]

    illuminated

    PROPOSED METHOD 3.1. Retinex-based Decomposition and Illumination Guid- ance According to Retinex theory [5, 6], a low-light image I ∈ RH×W×3 can be expressed as an element-wise product of a reflectance image R (same dimensions as I) and an illumina- tion map M ∈ RH×W : I = R ⊙ M. (1) In low-light conditions, M captures lighting variations, while R preser...

    work page

  5. [4]

    EXPERIMENTS 4.1. Real-world Low-light V-TIEE dataset We introduce the V-TIEE dataset, captured using a co-located visible (1440×1080px, 34° FOV) and thermal (640×512px, 24° HFOV) camera co-located using a gold dichroic mirror that transmits visible light and reflects thermal infrared. Final spatial alignment is refined using homography. The dataset includ...

    work page

  6. [5]

    Quantitative comparison of RT-X Net with baseline methods on LLVIP and V-TIEE dataset

    Self-Attention Network: An ablation study shows a significant performance drop when the cross-attention net- Method LLVIP real-world V-TIEE PSNR ↑ SSIM ↑ LPIPS ↓ SSIM ↑ KinD [23] 20.41 0.54 0.29 0.52 LLFlow [14] 21.53 0.57 0.27 0.56 LLFormer[15] 22.27 0.61 0.26 0.60 SNR Aware [7] 22.87 0.64 0.23 0.61 GSAD Net [8] 24.43 0.69 0.21 0.63 EnGAN [16] 25.67 0.74...

    work page

  7. [6]

    Another channel to concatenate Thermal Images: Thermal data is incorporated by concatenating thermal im- ages with RGB images to create a4-channel input, but this ap- proach shows limited performance gains. Adding a dedicated layer for thermal images provides marginal improvement but Input RGBInput ThermalEnGANLLFormerGSADRetinexformerRT-X NetGT/Reference...

    work page

  8. [7]

    The experimental results demonstrate that the proposed approach outperforms recent state-of-the-art methods by effectively incorporating ther- mal information

    CONCLUSION We present RT-X Net, a transformer network utilizing cross- attention to leverage thermal imagery for low-light enhance- ment under extreme conditions. The experimental results demonstrate that the proposed approach outperforms recent state-of-the-art methods by effectively incorporating ther- mal information. Ablation studies further validate ...

    work page

Show all 30 references
  1. [8]

    Degan: Mixed noise removal via generative adversarial net- works,

    Qiongshuai Lyu, Min Guo, and Zhao Pei, “Degan: Mixed noise removal via generative adversarial net- works,” Applied Soft Computing , vol. 95, pp. 106478, 2020

    2020

  2. [9]

    Deep multi-scale convolutional neural network for dy- namic scene deblurring,

    Seungjun Nah, Tae Hyun Kim, and Kyoung Mu Lee, “Deep multi-scale convolutional neural network for dy- namic scene deblurring,” in IEEE CVPR , 2017, pp. 3883–3891

    2017

  3. [10]

    Deblurgan: Blind motion deblurring using conditional adversarial net- works,

    Orest Kupyn, V olodymyr Budzan, Mykola Mykhailych, Dmytro Mishkin, and Ji ˇr´ı Matas, “Deblurgan: Blind motion deblurring using conditional adversarial net- works,” in IEEE CVPR, 2018, pp. 8183–8192

    2018

  4. [11]

    Learn- ing a single convolutional super-resolution network for multiple degradations,

    Kai Zhang, Wangmeng Zuo, and Lei Zhang, “Learn- ing a single convolutional super-resolution network for multiple degradations,” inIEEE CVPR, 2018, pp. 3262– 3271

    2018

  5. [12]

    Deep retinex decomposition for low-light en- hancement,

    Chen Wei, Wenjing Wang, Wenhan Yang, and Jiaying Liu, “Deep retinex decomposition for low-light en- hancement,” arXiv preprint arXiv:1808.04560, 2018

    2018 arXiv

  6. [13]

    Retinexformer: One-stage retinex-based transformer for low-light image enhance- ment,

    Yuanhao Cai, Hao Bian, Jing Lin, Haoqian Wang, Radu Timofte, and Yulun Zhang, “Retinexformer: One-stage retinex-based transformer for low-light image enhance- ment,” in IEEE/CVF ICCV, 2023, pp. 12504–12513

    2023

  7. [14]

    Snr-aware low-light image enhancement,

    Xiaogang Xu, Ruixing Wang, Chi-Wing Fu, and Jiaya Jia, “Snr-aware low-light image enhancement,” in IEEE/CVF CVPR, 2022, pp. 17714–17724

    2022

  8. [15]

    Global structure-aware dif- fusion process for low-light image enhancement,

    Jinhui Hou, Zhiyu Zhu, Junhui Hou, Hui Liu, Huan- qiang Zeng, and Hui Yuan, “Global structure-aware dif- fusion process for low-light image enhancement,” Ad- vances in Neural Information Processing Systems , vol. 36, 2024

    2024

  9. [16]

    On the role of thermal imaging in automotive applications: A critical review,

    Muhammad Ali Farooq, Waseem Shariff, David O’callaghan, Arcangelo Merla, and Peter Corcoran, “On the role of thermal imaging in automotive applications: A critical review,” IEEE Access, vol. 11, pp. 25152– 25173, 2023

    2023

  10. [17]

    Mfnet: Towards real-time semantic segmentation for autonomous vehi- cles with multi-spectral scenes,

    Qishen Ha, Kohei Watanabe, Takumi Karasawa, Yoshi- taka Ushiku, and Tatsuya Harada, “Mfnet: Towards real-time semantic segmentation for autonomous vehi- cles with multi-spectral scenes,” in 2017 IEEE/RSJ In- ternational Conference on Intelligent Robots and Sys- tems (IROS). IE...

    2017

  11. [18]

    Cmx: Cross- modal fusion for rgb-x semantic segmentation with transformers,

    Jiaming Zhang, Huayao Liu, Kailun Yang, Xinxin Hu, Ruiping Liu, and Rainer Stiefelhagen, “Cmx: Cross- modal fusion for rgb-x semantic segmentation with transformers,” IEEE Transactions on intelligent trans- portation systems, 2023

    2023

  12. [19]

    Contextual and vari- ational contrast enhancement,

    Turgay Celik and Tardi Tjahjadi, “Contextual and vari- ational contrast enhancement,” IEEE TIP, vol. 20, no. 12, pp. 3431–3441, 2011

    2011

  13. [21]

    Low-light image en- hancement with normalizing flow,

    Yufei Wang, Renjie Wan, Wenhan Yang, Haoliang Li, Lap-Pui Chau, and Alex Kot, “Low-light image en- hancement with normalizing flow,” in Proceedings of the AAAI conference on artificial intelligence , 2022, vol. 36, pp. 2604–2612

    2022

  14. [22]

    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 , 2023, vol. 37, pp. 2654–2662

    2023

  15. [23]

    Enlightengan: Deep light enhance- ment 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 enhance- ment without paired supervision,” IEEE transactions on image processing, vol. 30, pp. 2340–2349, 2021

    2021

  16. [24]

    Elegan: An efficient low light enhancement gan for unpaired supervision,

    Rohit Choudhary, T Harshith Reddy, and Mansi Sharma, “Elegan: An efficient low light enhancement gan for unpaired supervision,” in 2023 IEEE ICIP . IEEE, 2023, pp. 3105–3109

    2023

  17. [25]

    Rgb and lut based cross atten- tion network for image enhancement.,

    Tengfei Shi, Chenglizhao Chen, Yuanbo He, Wenfeng Song, and Aimin Hao, “Rgb and lut based cross atten- tion network for image enhancement.,” in BMVC, 2023, pp. 348–350

    2023

  18. [26]

    Rtfnet: Rgb- thermal fusion network for semantic segmentation of ur- ban scenes,

    Yuxiang Sun, Weixun Zuo, and Ming Liu, “Rtfnet: Rgb- thermal fusion network for semantic segmentation of ur- ban scenes,” IEEE RAL, vol. 4, no. 3, pp. 2576–2583, 2019

    2019

  19. [27]

    A deep thermal-guided approach for effective low-light visible image enhancement,

    Yanpeng Cao, Xi Tong, Fan Wang, Jiangxin Yang, Yan- long Cao, Sabin Tiberius Strat, and Christel-Loic Tisse, “A deep thermal-guided approach for effective low-light visible image enhancement,”Neurocomputing, vol. 522, pp. 129–141, 2023

    2023

  20. [28]

    Llvip: A visible-infrared paired dataset for low-light vision,

    Xinyu Jia, Chuang Zhu, Minzhen Li, Wenqi Tang, and Wenli Zhou, “Llvip: A visible-infrared paired dataset for low-light vision,” in IEEE/CVF ICCV , 2021, pp. 3496–3504

    2021

  21. [29]

    Noise-optimal capture for high dynamic range photography,

    Samuel W Hasinoff, Fr ´edo Durand, and William T Free- man, “Noise-optimal capture for high dynamic range photography,” in IEEE CVPR . IEEE, 2010, pp. 553– 560

    2010

  22. [30]

    Kin- dling the darkness: A practical low-light image en- hancer,

    Yonghua Zhang, Jiawan Zhang, and Xiaojie Guo, “Kin- dling the darkness: A practical low-light image en- hancer,” in ACM international conference on multime- dia, 2019, pp. 1632–1640

    2019

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.