Pith. sign in

REVIEW 5 major objections 5 minor 38 references

ADAM-Dehaze: Adaptive Density-Aware Multi-Stage Dehazing for Improved Object Detection in Foggy Conditions

T0 review · 5 major / 5 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 This paper argues that foggy images should be dehazed by a network selected according to the estimated fog density, and that doing so improves both restoration quality and downstream object detection compared with a single uniform…

desk verdict Plausible adaptive-dehazing idea, but the paper never actually tests the adaptive routing on real fog, and the abstract's numbers don't match the tables. read the letter →

arxiv 2506.15837 v1 pith:QMXXZEQB submitted 2025-06-16 cs.CV

Fatmah AlHindaassi , Mohammed Talha Alam , Fakhri Karray This is my paper
classification cs.CV
keywords adaptivedehazinghazedensityestimationobjectdetectioninfogatmosphericscatteringmodeldeepunfoldingnetworkmulti-stageimagerestorationfoggyscenesreal-world
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to show that fog removal is best done adaptively, with the network capacity and loss weighting matched to the estimated fog density. It proposes a framework that first computes a haze density score for each image, then routes the image to one of three dehazing networks designed for light, medium, or heavy fog. On synthetic Cityscapes and real-world RTTS images, the authors report PSNR gains of up to 2.1 dB, a 30% reduction in FADE, object detection mAP gains of up to 13 points, and a 20% reduction in inference time. If correct, this would mean that density-aware routing is a practical way to make dehazing genuinely useful for downstream tasks like autonomous driving.

What carries the argument

The load-bearing mechanism is the coupling between the Haze Density Estimation Network (HDEN) and the three CORUN branches. CORUN is a cooperative unfolding network that alternates steps estimating scene radiance and transmission; the paper instantiates it with 2, 4, and 6 stages, the heaviest branch adding transformer-based attention. HDEN's predicted score $d$ determines not only which branch runs, but also the mix of loss terms via $\gamma(d) \in \{0.3, 0.6, 0.9\}$. This coupling makes the network capacity and the optimization objective both vary with fog density, which is the backbone of the claimed gains.

What would settle it

On real foggy images with visibly dense fog (for example, RTTS images of heavy haze), record which branch HDEN selects. If most are sent to the Light branch, or if using the predicted routing gives no better detection or restoration than forcing all images through one fixed branch, the adaptive advantage collapses. A cleaner control on synthetic data is to compare the full system against an oracle router that uses the ground-truth beta value; a small gap would indicate HDEN contributes little.

Watch

Extended reading notes

Core claim

The paper claims to show that a density-aware, multi-branch system beats a monolithic dehazer on both image quality and detection. A lightweight Haze Density Estimation Network (HDEN) outputs a continuous score $d \in [0,1]$; thresholds $\alpha$ and $\beta$ route the input to a 2-stage CORUN-Light, 4-stage CORUN-Medium, or 6-stage CORUN-Complex branch. The same score sets the loss weight $\gamma$ between physical-coherence, perceptual, and density terms. On synthetic Cityscapes the authors report PSNR up to 23.95 dB, SSIM up to 0.9188, and LPIPS down to 0.0585, and on real-world RTTS they report FADE 0.828, BRISQUE 11.961, NIMA 5.346, and mean detection mAP 0.56, with detection improvements up to 13 mAP points over hazy inputs.

Load-bearing premise

The central assumption is that fog synthesized with the atmospheric scattering model at three fixed density values ($\beta = 0.03, 0.06, 0.09$) teaches the density estimator to route real-world foggy images correctly.

Editorial extensions

If this is right

  • In light fog the 2-stage branch runs at 18 ms per image, so the adaptive scheme is fast enough for real-time use in vehicles while reserving the 50 ms 6-stage branch for dense fog.
  • Because the dehazing branches are fine-tuned together with a detector, the reported mAP gains come from features aligned to detection, not only from cleaner pixels.
  • The density-modulated loss assigns higher weight to physical-model coherence in heavy fog, implying that physics-based constraints matter most precisely where pure learning struggles.
  • The synthetic FogIntensity-25K dataset with $\beta = 0.03, 0.06, 0.09$ gives the community paired hazy/clear images with depth-aware fog, enabling full-reference evaluation across intensity levels.

Reading between the lines

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

  • A direct test: compare predicted routing against oracle routing on real fog images. If the oracle does not beat the learned router, HDEN's classification accuracy is not what drives the reported gains.
  • The routing in Equation (4) is a hard switch; a soft mixture of branch outputs near the thresholds might behave better on continuous haze densities and on images that straddle two regimes.
  • Because HDEN is trained on three discrete synthetic beta values, real fog with spatially varying density and arbitrary atmospheric light creates a distribution gap; measuring the router's confidence on out-of-distribution fog would quantify this.
  • If the density-aware principle holds, the same router-plus-branch design could be transferred to other degradations such as rain, snow, or low light, each with its own cheap and expensive restoration paths.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The paper proposes ADAM-Dehaze, an adaptive dehazing framework that first classifies a foggy image into light, medium, or heavy fog using a Haze Density Estimation Network (HDEN), then routes the image through one of three CORUN-based dehazing branches, and finally feeds the restored image into YOLOv8n for object detection. The authors introduce a synthetic FogIntensity-25K dataset and a density-modulated loss combining physical-model coherence, perceptual, and density terms. The claimed contributions are improved restoration (up to 2.1 dB PSNR, 30% FADE reduction) and detection (up to 13 mAP points) on Cityscapes and RTTS, with reduced inference time.

Significance. If the central claims were fully supported, the idea of routing images by estimated fog density to specialized restoration branches, while co-optimizing with a downstream detector, would be a useful engineering contribution to fog-robust perception. The paper has several strengths: the authors commit to releasing code, models, and the FogIntensity-25K dataset; they include ablations for the loss components and the cooperative proximal mapping module; and they evaluate on a real-world benchmark (RTTS) in addition to synthetic Cityscapes data. However, the evidence as presented does not yet support the load-bearing claims. The adaptive routing mechanism, which is the paper's main novelty, is never evaluated as a complete system on real foggy images, and the headline quantitative gains in the abstract are not consistently reflected in the tables. The central hypothesis remains plausible but unverified.

major comments (5)
  1. [IV-E, Table V] The only real-world detection evaluation is performed using the CORUN-Light branch only, as stated in the table caption ('Our method, leveraging the CORUN-Light branch') and in Section IV-E. This means the full adaptive system (HDEN routing plus all three branches) is never tested on RTTS. The abstract's claim of 'up to 13 points' mAP improvement is not visible in Table V: the mean mAP is 0.56 versus RIDCP's 0.55, a 1-point difference, and no no-dehazing YOLOv8n baseline is reported. The authors should report results for the full adaptive pipeline on RTTS, the routing distribution over the three branches, and a comparison against each fixed branch, in addition to the fixed-branch baseline.
  2. [IV-C-1, Table IV] Table IV reports SSIM, PSNR, and LPIPS only for the proposed method, with no baseline rows, so the claimed 'up to 2.1 dB' PSNR improvement cannot be confirmed from the presented data. The text states that ADAM-Dehaze 'consistently achieves the lowest distortion scores and highest perceptual/structural metrics,' but no comparison values are shown. The authors should add the same metrics for DAD, MBDN, RIDCP, and PSD in Table IV, matching the comparison columns in Table III.
  3. [III-B-2, Eq. (4), Algorithm 1] The routing thresholds α and β in Eq. (4) and Algorithm 1 are never specified, and the HDEN classification accuracy (99.80%, Table II) is measured only on 5,000 synthetic test images. Without explicit threshold values and without any real-world routing accuracy evaluation, the behavior of the adaptive system on RTTS is undefined. The authors should provide the chosen α and β values and report HDEN accuracy or routing statistics on real foggy images, including which branches RTTS images are routed to.
  4. [IV-F, Table VI] The inference-time claims are inconsistent across the paper. The abstract states a 20% inference-time reduction; Section IV-F states a 33% average latency reduction and 23% FLOP savings; and the text below Table VI says '20–40%.' From Table VI, the adaptive average time (30 ms) versus the fixed 6-stage model (45 ms) gives a 33% reduction, not 20%. The authors should reconcile these numbers and report the actual measured routing distribution used to compute the average.
  5. [III-B-3, Eq. (8)] The density loss Ldens = ||D_theta_D(Jout)||_1 uses the same HDEN network that performs routing, but the paper does not state whether HDEN is frozen or fine-tuned during the joint training. If D_theta_D is optimized jointly with this loss, a trivial solution is for the density estimator to output zero on all dehazed images, making the loss vacuous without enforcing real haze removal. The training status of HDEN and any stop-gradient or fixed-weight treatment should be stated explicitly.
minor comments (5)
  1. [Section II, Related Work] The text attributes CORUN to reference [12], but [12] is the MSBDN paper by Dong et al. If CORUN is a separate method, it should have its own citation; if not, the description of 'cooperative unfolding modules' needs a correct reference.
  2. [Section IV-C-1, Eq. (6)] The coherence loss in Eq. (6) implicitly assumes global atmospheric light A = 1. This assumption is not stated in the text and is part of the physical-model coherence claim; it should be made explicit.
  3. [Abstract and Section IV-C-2] The abstract claims a 30% FADE reduction, but Table V shows FADE of 0.828 for the proposed method versus 0.944 for RIDCP, which is a 12.3% reduction, and the text itself says '12.7% in FADE.' The 30% figure appears only in the abstract and in Section VI ('up to 30% lower FADE'). These numbers should be reconciled.
  4. [Table VII and Table V] The ablation table reports mAP of 75.0% for the full model, while Table V reports mean mAP of 0.56 (56%) on RTTS. The differing scales and datasets are not explained; the ablation should state which dataset and protocol produce the 75.0% value.
  5. [IV-E, Section V] The claim in Section IV-E that 'these improvements are consistent across the different object categories' is only partially supported: the Bus category shows 0.31 for Ours versus 0.32 for RIDCP, so one category is actually lower. The wording should be softened or the per-category results discussed more carefully.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation; the central claims rest on external benchmarks and the only self-referential term is a non-load-bearing auxiliary loss.

full rationale

The claimed chain—HDEN score, branch routing, CORUN restoration, YOLOv8 detection—is not equivalent to its inputs by construction. HDEN is trained on synthetic fog labels derived from the atmospheric scattering model with beta in {0.03, 0.06, 0.09}; the dehazing branches are supervised by reconstruction and perceptual losses; detection is fine-tuned on FogIntensity-25K. None of the headline metrics (PSNR/SSIM/LPIPS, FADE/BRISQUE/NIMA, YOLOv8 mAP) is computed from HDEN's own score: PSNR/SSIM use Cityscapes ground truth, FADE is the external heuristic of Choi et al., and mAP is measured with YOLOv8n. The one self-referential term, Ldens = ||D_theta_D(Jout)||_1 in Eq. (8), merely adds an auxiliary penalty; ablations show removing it costs about 0.8 dB and 1.2 mAP, so the main result does not reduce to it. Reference [32] (FLARE) is authored by two of the present authors but appears only in related-work discussion, not as a load-bearing uniqueness or existence theorem. The RTTS evaluation uses only the CORUN-Light branch and the thresholds alpha and beta are never specified; these are missing-evidence or support problems that weaken the adaptive-routing claim, but they are not circular reasoning. Overall, no step in the paper's derivation chain is identical to its inputs.

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

The framework rests on a hand-constructed synthetic fog distribution (beta values), an unstated set of routing thresholds, and a hand-tuned loss schedule. The central empirical claim also assumes that synthetic-to-real transfer works at both the density-estimation and detection stages.

free parameters (4)
  • Routing thresholds alpha and beta
    Used in Eq. 4 and Eq. 5 to select branch and set loss weight gamma. Values are never reported, so they are chosen by hand or fit to the synthetic validation data, and they directly control the adaptive behavior that is the central contribution.
  • Loss weights gamma = 0.3, 0.6, 0.9
    Hand-picked weights for the coherence-perceptual tradeoff in Eq. 5, dependent on the same unstated alpha/beta thresholds.
  • Synthetic fog strengths beta = 0.03, 0.06, 0.09
    Hand-selected scattering coefficient values used to generate the FogIntensity-25K dataset and to define the three fog classes. These values are inputs, not outcomes, but they are chosen by the authors and determine the taxonomy of fog levels.
  • Perceptual loss layer weights tau_i
    Weights in Eq. 7 for VGG-19 feature layers are not specified; they are conventional but are free numerical choices that affect training.
assumptions (4)
  • domain assumption The atmospheric scattering model P = J * t + A(1 - t) with t = exp(-beta*d) is an adequate generative model for real-world fog.
    Used throughout Section III-A to create synthetic hazy pairs from Cityscapes and Synscapes depth maps. If real fog deviates substantially from this model, the synthetic-to-real transfer fails.
  • ad hoc to paper The attenuation coefficient beta is constant per scene and the global atmospheric light A is uniform (implicitly set to 1 in the coherence loss Eq. 6).
    The coherence loss reconstructs the hazy image as J_out * T_out + (1 - T_out), assuming A = 1 and spatially uniform light, without discussing color or non-uniform illumination.
  • domain assumption HDEN's density score, learned on synthetic labels, is an ordinal measure of real-world fog density.
    The routing in Eq. 4 depends on this score; the paper reports classification accuracy only on a synthetic test set and does not validate the score on RTTS images.
  • domain assumption YOLOv8n fine-tuned on synthetic foggy images transfers to real-world RTTS detection.
    The detection improvement is measured after fine-tuning on FogIntensity-25K; no comparison against the detector applied directly to hazy images is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ADAM-Dehaze: Adaptive Density-Aware Multi-Stage Dehazing for Improved Object Detection in Foggy Conditions." pith.science (2026). https://pith.science/paper/QMXXZEQB

@misc{pith2026250615837,
  author       = {Pith},
  title        = {Pith review of: ADAM-Dehaze: Adaptive Density-Aware Multi-Stage Dehazing for Improved Object Detection in Foggy Conditions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QMXXZEQB}},
  note         = {Machine review of arXiv:2506.15837}
}
read the original abstract

Adverse weather conditions, particularly fog, pose a significant challenge to autonomous vehicles, surveillance systems, and other safety-critical applications by severely degrading visual information. We introduce ADAM-Dehaze, an adaptive, density-aware dehazing framework that jointly optimizes image restoration and object detection under varying fog intensities. A lightweight Haze Density Estimation Network (HDEN) classifies each input as light, medium, or heavy fog. Based on this score, the system dynamically routes the image through one of three CORUN branches: Light, Medium, or Complex, each tailored to its haze regime. A novel adaptive loss balances physical-model coherence and perceptual fidelity, ensuring both accurate defogging and preservation of fine details. On Cityscapes and the real-world RTTS benchmark, ADAM-Dehaze improves PSNR by up to 2.1 dB, reduces FADE by 30 percent, and increases object detection mAP by up to 13 points, while cutting inference time by 20 percent. These results highlight the importance of intensity-specific processing and seamless integration with downstream vision tasks. Code available at: https://github.com/talha-alam/ADAM-Dehaze.

Figures

Figures reproduced from arXiv: 2506.15837 by the authors.

Figure 1
Figure 1. Conceptual illustration of ADAM-Dehaze. Given [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Architecture of the proposed ADAM-Dehaze framework. The process begins with generating foggy images [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Comparison of clear weather images from Syn [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Object detection under complex haze using PSD [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 4
Figure 4. Figure 4: Qualitative comparison of dehazing performance [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

38 extracted references · 33 canonical work pages

  1. [1]

    Vision and the atmo- sphere,

    S. G. Narasimhan and S. K. Nayar, “Vision and the atmo- sphere,” International journal of computer vision, vol. 48, pp. 233–254, 2002

    work page 2002

  2. [2]

    Single image haze removal using dark channel prior,

    K. He, J. Sun, and X. Tang, “Single image haze removal using dark channel prior,” IEEE transactions on pattern analysis and machine intelligence, vol. 33, no. 12, pp. 2341–2353, 2010

    work page 2010

  3. [3]

    Benchmarking single-image dehazing and beyond,

    B. Li, W. Ren, D. Fu, D. Tao, D. Feng, W. Zeng, and Z. Wang, “Benchmarking single-image dehazing and beyond,” IEEE Transactions on Image Processing, vol. 28, no. 1, pp. 492–505, 2018

    work page 2018

  4. [4]

    Benchmark- ing robustness in object detection: Autonomous driving when winter is coming,

    C. Michaelis, B. Mitzkus, R. Geirhos, E. Rusak, O. Bring- mann, A. S. Ecker, M. Bethge, and W. Brendel, “Benchmark- ing robustness in object detection: Autonomous driving when winter is coming,”arXiv preprint arXiv:1907.07484, 2019

    arXiv 1907

  5. [5]

    Yolov3: An incremental improve- ment,

    J. Redmon and A. Farhadi, “Yolov3: An incremental improve- ment,” arXiv preprint arXiv:1804.02767, 2018

    arXiv 2018

  6. [6]

    Faster r-cnn: Towards real-time object detection with region proposal networks,

    S. Ren, K. He, R. Girshick, and J. Sun, “Faster r-cnn: Towards real-time object detection with region proposal networks,” in Advances in Neural Information Processing Systems (NIPS), 2015, pp. 91–99

    work page 2015

  7. [7]

    Model adaptation with synthetic and real data for semantic dense foggy scene understanding,

    C. Sakaridis, D. Dai, S. Hecker, and L. Van Gool, “Model adaptation with synthetic and real data for semantic dense foggy scene understanding,” in Proceedings of the european conference on computer vision (ECCV), 2018, pp. 687–704

    work page 2018

  8. [8]

    Semanticunderstandingoffoggysceneswithpurelysynthetic data,

    M.Hahner,D.Dai,C.Sakaridis,J.-N.Zaech,andL.VanGool, “Semanticunderstandingoffoggysceneswithpurelysynthetic data,” in2019 IEEE Intelligent Transportation Systems Con- ference (ITSC). IEEE, 2019, pp. 3675–3681

    work page 2019

Show all 38 references
  1. [9]

    A re- view on dark channel prior based image dehazing algorithms,

    S. Lee, S. Yun, J.-H. Nam, C. S. Won, and S.-W. Jung, “A re- view on dark channel prior based image dehazing algorithms,” EURASIP Journal on Image and Video Processing, vol. 2016, pp. 1–23, 2016

    2016

  2. [10]

    Saliency-based dark chan- nel prior model for single image haze removal,

    L. Zhang, S. Wang, and X. Wang, “Saliency-based dark chan- nel prior model for single image haze removal,”IET Image processing, vol. 12, no. 6, pp. 1049–1055, 2018

    2018

  3. [11]

    Haze visibil- ity enhancement: A survey and quantitative benchmarking,

    Y. Li, S. You, M. S. Brown, and R. T. Tan, “Haze visibil- ity enhancement: A survey and quantitative benchmarking,” Computer Vision and Image Understanding, vol. 165, pp. 1– 16, 2017

    2017

  4. [12]

    Multi-scale boosted dehazing network with dense feature fusion,

    H. Dong, J. Pan, L. Xiang, Z. Hu, X. Zhang, F. Wang, and M.- H. Yang, “Multi-scale boosted dehazing network with dense feature fusion,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020, pp. 2157–2167

    2020

  5. [13]

    Gated fusion network for single image dehazing,

    W. Ren, L. Ma, J. Zhang, J. Pan, X. Cao, W. Liu, and M.-H. Yang, “Gated fusion network for single image dehazing,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 3253–3261

    2018

  6. [14]

    Aod-net: All-in-one dehazing network,

    B. Li, X. Peng, Z. Wang, J. Xu, and D. Feng, “Aod-net: All-in-one dehazing network,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2017, pp. 4770–4778

    2017

  7. [15]

    End-to-end united video dehazing and detection,

    ——, “End-to-end united video dehazing and detection,” in Proceedings of the AAAI conference on artificial intelligence, vol. 32, no. 1, 2018

    2018

  8. [16]

    Unified density-aware image dehazing and object detection in real- world hazy scenes,

    Z. Zhang, L. Zhao, Y. Liu, S. Zhang, and J. Yang, “Unified density-aware image dehazing and object detection in real- world hazy scenes,” inProceedings of the Asian Conference on Computer Vision, 2020

    2020

  9. [17]

    Perceiving and modeling density for image dehaz- ing,

    T. Ye, Y. Zhang, M. Jiang, L. Chen, Y. Liu, S. Chen, and E. Chen, “Perceiving and modeling density for image dehaz- ing,” inEuropean Conference on Computer Vision. Springer, 2022, pp. 130–145

    2022

  10. [18]

    Ridcp: Revitalizing real image dehazing via high-quality codebook priors,

    R.-Q. Wu, Z.-P. Duan, C.-L. Guo, Z. Chai, and C. Li, “Ridcp: Revitalizing real image dehazing via high-quality codebook priors,” inProceedings of the IEEE/CVF conference on com- puter vision and pattern recognition, 2023, pp. 22282–22291

    2023

  11. [19]

    Psd: Principled synthetic-to-real dehazing guided by physical priors,

    Z. Chen, Y. Wang, Y. Yang, and D. Liu, “Psd: Principled synthetic-to-real dehazing guided by physical priors,” inPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 7180–7189

    2021

  12. [20]

    Self-augmented unpaired image dehazing via density and depth decomposition,

    Y. Yang, C. Wang, R. Liu, L. Zhang, X. Guo, and D. Tao, “Self-augmented unpaired image dehazing via density and depth decomposition,” inProceedings of the IEEE/CVF con- ference on computer vision and pattern recognition, 2022, pp. 2037–2046

    2022

  13. [21]

    Deep generalized unfold- ing networks for image restoration,

    C. Mou, Q. Wang, and J. Zhang, “Deep generalized unfold- ing networks for image restoration,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recog- nition, 2022, pp. 17399–17410

    2022

  14. [22]

    Ffa-net: Featurefusionattentionnetworkforsingleimagedehazing,

    X. Qin, Z. Wang, Y. Bai, X. Xie, and H. Jia, “Ffa-net: Featurefusionattentionnetworkforsingleimagedehazing,”in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 34, no. 07, 2020

    2020

  15. [23]

    The cityscapes dataset for semantic urban scene understanding,

    M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Benenson, U. Franke, S. Roth, and B. Schiele, “The cityscapes dataset for semantic urban scene understanding,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 3213–3223

    2016

  16. [24]

    An enhanced haze re- moval: Using dcp and enriched-invariant features,

    R. Tanya, F. Akhtar, and Mahipal, “An enhanced haze re- moval: Using dcp and enriched-invariant features,” inInter- national Conference on Micro-Electronics and Telecommuni- cation Engineering. Springer, 2021, pp. 257–267

    2021

  17. [25]

    Single image dehazing with bright object handling,

    I. Riaz, X. Fan, and H. Shin, “Single image dehazing with bright object handling,”IET Computer Vision, vol. 10, no. 8, pp. 817–827, 2016

    2016

  18. [26]

    Phdnet: A novel physic-aware dehazing network for remote sensing images,

    Z. Lihe, J. He, Q. Yuan, X. Jin, Y. Xiao, and L. Zhang, “Phdnet: A novel physic-aware dehazing network for remote sensing images,” Information Fusion, vol. 106, p. 102277, 2024

    2024

  19. [27]

    End to end system for hazy image classification and recon- struction based on mean channel prior using deep learning network,

    S. Satrasupalli, E. Daniel, S. R. Guntur, and S. Shehanaz, “End to end system for hazy image classification and recon- struction based on mean channel prior using deep learning network,” IET Image Processing, vol. 14, no. 17, pp. 4736– 4743, 2020

    2020

  20. [28]

    Dehazenet: An end-to-end system for single image haze removal,

    B. Cai, X. Xu, K. Jia, C. Qing, and D. Tao, “Dehazenet: An end-to-end system for single image haze removal,”IEEE transactions on image processing, vol. 25, no. 11, pp. 5187– 5198, 2016

    2016

  21. [29]

    Griddehazenet: Attention-based multi-scale network for image dehazing,

    X. Liu, Y. Ma, Z. Shi, and J. Chen, “Griddehazenet: Attention-based multi-scale network for image dehazing,” in Proceedings of the IEEE International Conference on Com- puter Vision (ICCV), 2019, pp. 7314–7323

    2019

  22. [30]

    Acceler- ated unfolding network for medical image reconstruction with efficient information flow,

    M. Qin, R. Yang, M. Wan, C. Zhang, and T. Li, “Acceler- ated unfolding network for medical image reconstruction with efficient information flow,” in International Conference on Intelligent Robotics and Applications. Springer, 2023, pp. 43–54

    2023

  23. [31]

    Optimiz- ing wireless networks with deep unfolding: Comparative study on two deep unfolding mechanisms,

    A. Adam, M. A. Elhassan, and E. M. Diallo, “Optimiz- ing wireless networks with deep unfolding: Comparative study on two deep unfolding mechanisms,” arXiv preprint arXiv:2403.18930, 2024

    2024 arXiv

  24. [32]

    Flare up yourdata:Diffusion-basedaugmentationmethodinastronom- ical imaging,

    M. T. Alam, R. Imam, M. Guizani, and F. Karray, “Flare up yourdata:Diffusion-basedaugmentationmethodinastronom- ical imaging,”arXiv preprint arXiv:2405.13267, 2024

    2024 arXiv

  25. [33]

    Ssd: Single shot multibox detector,

    W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y. Fu, and A. C. Berg, “Ssd: Single shot multibox detector,” in European Conference on Computer Vision (ECCV), 2016, pp. 21–37

    2016

  26. [34]

    Referenceless prediction of perceptual fog density and perceptual image defogging,

    L. K. Choi, J. You, and A. C. Bovik, “Referenceless prediction of perceptual fog density and perceptual image defogging,” IEEE Transactions on Image Processing, vol. 24, no. 11, pp. 3888–3901, 2015

    2015

  27. [35]

    No-reference image quality assessment in the spatial domain,

    A. Mittal, A. K. Moorthy, and A. C. Bovik, “No-reference image quality assessment in the spatial domain,”IEEE Trans- actions on image processing, vol. 21, no. 12, pp. 4695–4708, 2012

    2012

  28. [36]

    Nima: Neural image assessment,

    H. Talebi and P. Milanfar, “Nima: Neural image assessment,” IEEE Transactions on Image Processing, vol. 27, no. 8, pp. 3998–4011, 2018

    2018

  29. [37]

    Synscapes: A photorealistic synthetic dataset for street scene parsing,

    M. Wrenninge and J. Unger, “Synscapes: A photorealistic synthetic dataset for street scene parsing,” arXiv preprint arXiv:1810.08705, 2018

    2018 arXiv

  30. [38]

    Domain adap- tation for image dehazing,

    Y. Shao, L. Li, W. Ren, C. Gao, and N. Sang, “Domain adap- tation for image dehazing,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020, pp. 2808–2817

    2020

Pith tools

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