Pith. sign in

REVIEW 5 major objections 6 minor 66 references

MetaShadow: Object-Centered Shadow Detection, Removal, and Synthesis

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

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

Pith's one-line read MetaShadow claims that one object-mask-conditioned model can detect, remove, and re-synthesize shadows, and that sharing shadow-rich features between a GAN-based analyzer and a diffusion-based synthesizer is what makes all three tasks…

desk verdict A solid three-in-one shadow editing system with real gains on detection and removal, but the synthesis mechanism needs a cleaner ablation and the detection comparison isn't like-for-like. read the letter →

arxiv 2412.02635 v1 pith:LL2FNJ5Q submitted 2024-12-03 cs.CV

Tianyu Wang , Jianming Zhang , Haitian Zheng , Zhihong Ding , Scott Cohen , Zhe Lin , Wei Xiong , Chi-Wing Fu
show 2 more authors
Luis Figueroa Soo Ye Kim
This is my paper · ORCID
classification cs.CV
keywords object-centeredshadowdetectionremovalsynthesisknowledgetransferGANanddiffusionmodelsimageeditingreference-basedMOSdataset
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 show that shadow detection, removal, and synthesis do not have to be three separate models: one object-centered framework, MetaShadow, can handle all three, and the three tasks get better because they share knowledge. The framework pairs a GAN-based Shadow Analyzer, which takes an image and an object mask and produces both the object's shadow mask and a shadow-free image, with a reference-based diffusion Shadow Synthesizer that draws a new shadow. The shared knowledge is carried by multi-scale intermediate features from the Analyzer, injected into the Synthesizer through cross-attention after resizing to $32\times32$. The reported results are concrete: detection mIoU on SOBA rises from 55.8 to 71.0, removal bbox PSNR on DESOBA beats ShadowDiffusion by 8.7 dB, and synthesis Local RMSE on Video DESOBA falls from 51.73 to 36.54. A sympathetic reader would take the paper's central assertion to be that object-centered editing becomes far more realistic when shadows are treated as an editable property of the object, not as an afterthought.

What carries the argument

The load-bearing object is the shadow-knowledge-transfer channel between the two stages. From the Shadow Analyzer's spatial decoder, multi-scale features $F^i_{ms}$ at resolutions 16 through 128 are resized to a common $32\times32$ grid, concatenated into a tensor of shape $[N,1348,32,32]$, and passed through an adaptor $T(\cdot)$ made of a 2D convolution followed by a 1D convolution; an MLP then lifts the embedding dimension from 1344 to 2048, yielding the shadow embedding $E_s$ of shape $[N,1024,2048]$ that is fed into the diffusion U-Net's cross-attention layers as key and value. The paper's argument is that these features carry exactly the shadow properties the synthesizer needs—position, softness, intensity, color, direction—so the U-Net learns to paint a shadow that matches the reference instead of inventing one from text or generic image embeddings.

What would settle it

Replace the Shadow Analyzer's transferred features with features taken from the object's shadow-free region or from a different object while keeping everything else fixed; if synthesis Local RMSE does not degrade sharply, the paper's claim that these features carry transferable shadow knowledge is wrong.

Watch

Extended reading notes

Core claim

The core discovery, on the paper's own terms, is that a single framework can jointly perform object-centered shadow detection, removal, and synthesis, and that the joint training actually improves each task rather than compromising it. The Shadow Analyzer, built on an inpainting GAN, detects the shadow cast by a masked object and removes it; the Shadow Synthesizer, a reference-based diffusion model, then generates a new shadow for the object at a target location. The mechanism that makes the combination work is shadow knowledge transfer: the Analyzer's shadow-rich multi-scale features, after being resized to a uniform $32\times32$ resolution, are injected into the Synthesizer's cross-attention layers as a shadow embedding. The paper argues that this beats both pure GAN synthesis, which struggles with plausible shadow shapes, and text- or CLIP-conditioned diffusion synthesis, which lacks fine-grained control over shadow intensity, softness, color, and direction. It also claims a practical bonus: the conditioned synthesizer needs only four denoising steps, against 30 for the DDPM-based baseline and 50 for SGDiffusion.

Load-bearing premise

The framework depends on the assumption that the $32\times32$ multi-scale features injected into the synthesizer encode the shadow's transferable properties—position, softness, intensity, color, and direction—rather than the texture or color of the specific object or background; if that assumption fails, the claimed synthesis gains would not carry over to unseen objects and scenes.

Editorial extensions

If this is right

  • A single object-mask-conditioned model can produce the object's shadow mask, a shadow-free version of the image, and a re-synthesized shadow, so editing pipelines no longer need to stitch together separate detectors and removers.
  • Object-centered shadow detection on SOBA improves from 55.8 to 71.0 mIoU, meaning per-object shadow masks become reliable enough for downstream editing.
  • Shadow removal on DESOBA reaches 32.97 dB bbox PSNR, an 8.7 dB gain over ShadowDiffusion even when that baseline is given ground-truth masks, implying cleaner shadow-free regions.
  • Shadow synthesis reduces Local RMSE to 36.84 on DESOBA and 36.54 on Video DESOBA, and does so in four diffusion steps, so reference-based shadow synthesis becomes both more accurate and much faster.
  • Because the synthesizer is conditioned on a reference shadow, users can control shadow color, softness, and direction by choosing the reference, which supports creative edits like relocating objects with matched lighting.

Reading between the lines

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

  • If the transfer mechanism generalizes, then any object-conditional GAN that produces shadow-aware features could similarly condition a diffusion model, turning shadow knowledge transfer into a general recipe rather than a task-specific trick.
  • The four-step inference advantage suggests the feature conditioning may sharply reduce the diffusion sampling budget, which could matter for interactive editing tools even though the paper does not systematically study sampling efficiency.
  • A testable extension is to probe the transferred features under out-of-distribution lighting, such as indoor versus outdoor scenes or colored lights, to see whether softness and direction generalize beyond the static-background tripod videos in Video DESOBA.
  • The construction of a synthetic MOS dataset plus two real evaluation sets implies a reusable data recipe: render moving objects with shadows in Blender, then evaluate on real videos; the same recipe could apply to other intrinsic image decomposition tasks such as reflection or illumination editing.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The manuscript introduces MetaShadow, a two-stage framework that combines an object-centered GAN (Shadow Analyzer) for shadow detection and removal with a diffusion-based Shadow Synthesizer for reference-based shadow synthesis. The central idea is to transfer multi-scale intermediate features of the Shadow Analyzer into the diffusion model's cross-attention layers, providing shadow knowledge (position, softness, intensity, color, direction) to guide synthesis. The authors also introduce a synthetic training set (MOS), two real-world test sets (Moving DESOBA, Video DESOBA), and report large improvements over existing baselines on detection (SOBA mIoU 71.0 vs 55.8), removal (Bbox PSNR 32.97 vs 24.28), and synthesis (DESOBA Local RMSE 36.84 vs 56.44 for SGRNet). The paper claims to be the first framework to jointly handle all three object-centered shadow tasks with a shared-knowledge transfer mechanism.

Significance. If the claims are supported, this is a meaningful advance for object-centered image editing: a single framework that detects, removes, and synthesizes object shadows could simplify editing workflows and improve compositing realism. The proposed mechanism of injecting GAN-derived features into a diffusion model is a plausible and interesting approach to controllable shadow synthesis, and the new datasets are likely to be useful to the community. The paper is clearly written, includes visual comparisons and an ablation study, and the authors state that they will release the new test sets. However, several load-bearing issues in the evaluation protocol and in the evidence for the knowledge-transfer mechanism currently prevent full confidence in the SOTA claims.

major comments (5)
  1. [Sec. 5.1, Table 2] The shadow detection comparison is not apples-to-apples. The table caption notes that SSISv2 must automatically detect shadow-object instance pairs, whereas MetaShadow receives the object mask as input. The mIoU gain from 55.8 to 71.0 can therefore be substantially attributed to the oracle object mask rather than to superior shadow detection. To support the claimed SOTA detection performance, the authors should either evaluate MetaShadow with predicted object masks from an off-the-shelf object detector, or also report object detection accuracy so that both methods solve the same task under the same input protocol.
  2. [Sec. 4.2 and Fig. 3] The reference-based synthesis setup is not precisely specified. The text describes the Synthesizer as taking an object mask and an image containing the moved object, while the reference shadow appears to enter only through the Shadow Analyzer's features Fms computed from the original input image. For DESOBA the reference is described as another object in the image, and for Video DESOBA the first frame is used as the reference, but the paper does not state which image is fed to Stage I in these cases or how the network distinguishes the reference shadow from the target object's original shadow. Please provide a precise, implementable description of the reference input and the information flow, and clarify how a separate reference object or shadow is represented.
  3. [Sec. 5.2, Table 5] The ablation for the shadow knowledge transfer does not rule out appearance leakage. Fms includes features from resolutions 16 to 128, resized to 32x32, and the paper itself notes in Sec. 5.2 that larger-resolution features contain texture information within the shadow region. The ablation replaces Fms with T5 text or CLIP image embeddings, but it does not control for the amount of pixel-level image content injected or test whether the benefit survives when object appearance is changed while shadow properties are held fixed. Without such a control experiment, the claim that Fms encodes transferable shadow properties rather than object texture is not established, and the synthesis gains may not generalize to unseen objects.
  4. [Sec. 5.1, Table 4] The SOTA synthesis claim is metric-dependent. On Video DESOBA, MetaShadow's Bbox PSNR (23.14) is far below SGDiffusion (36.54), while on DESOBA the Bbox SSIM gain over SGRNet is small (93.56 vs 91.08). The abstract highlights the reduction in local RMSE from 51.73 to 36.54, but this selects the most favorable dataset-level metric. The paper should report all metrics with error bars or significance tests, and discuss explicitly why local RMSE is the appropriate primary metric, rather than presenting the result as uniformly superior.
  5. [Sec. 5.1 and Table 1] ObjectDrop [51] is listed in Table 1 as supporting detection, removal, and synthesis in an object-centered, reference-based manner, yet it is not included in any quantitative comparison. Since ObjectDrop is a very recent joint framework that appears to be highly relevant, omitting it weakens the claim that no existing method handles the same combination of tasks and that MetaShadow achieves SOTA. The authors should compare against ObjectDrop on the available benchmarks or justify its exclusion.
minor comments (6)
  1. [Sec. 3] The Moving DESOBA and Video DESOBA test sets are introduced without basic statistics (number of images, number of objects, resolution, annotation type). Please add a summary table in the main text or supplementary material.
  2. [Sec. 4.2] The adaptor T(·) is described as containing 'a 2D convolution layer followed by a 1D convolution layer'; this is likely intended as a 1x1 convolution, but the text should be clarified or corrected.
  3. [Eq. (1)] The notation Mes is used in the loss but not defined in the main text. Please state what Mes represents, when it is provided as input, and how it relates to the optional shadow mask in the Synthesizer.
  4. [Sec. 5.1, Table 3 caption] The caption should explicitly state that MetaShadow takes an object mask as input while ShadowDiffusion takes a shadow mask, since the protocol difference is essential for interpreting the comparison.
  5. [Sec. 5.2] The four-step diffusion inference setting is mentioned only in the final paragraph of Sec. 5.2. Please state the inference step count in Sec. 4.2 or the implementation details, and report the step counts used for all baselines for transparency.
  6. [References] References [47] and [48] appear to describe the same paper (Instance Shadow Detection, CVPR 2020) with different metadata; please consolidate to avoid a duplicate citation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the paper's claims are empirical, evaluated on held-out benchmarks, with no fitted parameter relabeled as a prediction and no load-bearing self-citation chain.

full rationale

MetaShadow's central claims are empirical and are assessed on held-out test sets (SOBA, DESOBA test set, Moving DESOBA, Video DESOBA) against external baselines. The framework is trained with standard supervised losses; no parameter-free derivation is claimed, and no fitted parameter is renamed as a prediction. The shadow-knowledge transfer is implemented by injecting Shadow Analyzer features into the diffusion synthesizer and is learned jointly, which is a conditioning design rather than a circular step. The paper's self-citations (e.g., prior instance shadow detection work and dataset SOBA) are used as prior art or baselines, not as the justification for the central claim. The paper itself notes in Sec. 5.2 that larger-resolution features contain unwanted texture information, which the authors acknowledge as a design tradeoff; this is a soundness or generalization concern, not circularity. The ablation in Table 5 compares against T5-text and CLIP-image conditions and supports the empirical claim that task-specific features help, though it does not fully rule out appearance leakage; again, this is an experimental-control limitation, not a case where the result reduces by construction to its inputs. No derived quantity is equivalent to an input by definition, and no uniqueness theorem or self-citation is invoked to force the architecture choice. The honest finding is therefore no significant circularity.

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

The ledger lists the hand-picked training and inference settings, the domain assumptions about input availability, and the design assumption that GAN features transfer shadow knowledge. No new physical entities are introduced; the model components are learned systems, and the new datasets are artifacts rather than postulated entities.

free parameters (5)
  • Diffusion inference steps = 4
    Chosen by hand for speed and quality; the paper states all results use a four-step setting, not derived from data.
  • Shadow embedding dimension = 2048
    The adaptor expands 1344 to 2048 dimensions via an MLP; selected without reported search.
  • Multi-scale feature resize size = 32x32
    Features from 16 to 128 resolutions are resized to uniform 32x32 before injection; chosen after observing texture leakage at larger resolutions.
  • Training resolutions = 512x512 (Analyzer), 128x128 (Synthesizer)
    Inference resolutions chosen by hand; synthesis results are upsampled to 256x256 for comparison.
  • DESOBA oversampling multiplier = 10 iterations
    DESOBA is iterated 10 times to balance dataset sizes; chosen without reported sensitivity analysis.
assumptions (4)
  • domain assumption Ground-truth object masks are available at test time as inputs for detection and removal.
    The task definition and Table 2 comparison rely on this; the claim of superior detection does not cover autonomous object-shadow association.
  • domain assumption A reference object-shadow pair or the object's original shadow is available for synthesis.
    Section 4 states 'A reference object casting a shadow is often available'; if absent, the method's controllability and quality are not characterized.
  • ad hoc to paper Blender-rendered MOS data is sufficiently realistic to transfer to natural images.
    Training uses synthetic MOS plus real datasets; the paper provides no cross-domain analysis separating synthetic from real contribution.
  • ad hoc to paper Shadow Analyzer intermediate features encode shadow properties rather than object texture at the chosen scales.
    Section 5.2 observes texture at larger resolutions and therefore restricts to 32x32; this is a design assumption validated only by ablation on one dataset.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MetaShadow: Object-Centered Shadow Detection, Removal, and Synthesis." pith.science (2026). https://pith.science/paper/LL2FNJ5Q

@misc{pith2026241202635,
  author       = {Pith},
  title        = {Pith review of: MetaShadow: Object-Centered Shadow Detection, Removal, and Synthesis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LL2FNJ5Q}},
  note         = {Machine review of arXiv:2412.02635}
}
read the original abstract

Shadows are often under-considered or even ignored in image editing applications, limiting the realism of the edited results. In this paper, we introduce MetaShadow, a three-in-one versatile framework that enables detection, removal, and controllable synthesis of shadows in natural images in an object-centered fashion. MetaShadow combines the strengths of two cooperative components: Shadow Analyzer, for object-centered shadow detection and removal, and Shadow Synthesizer, for reference-based controllable shadow synthesis. Notably, we optimize the learning of the intermediate features from Shadow Analyzer to guide Shadow Synthesizer to generate more realistic shadows that blend seamlessly with the scene. Extensive evaluations on multiple shadow benchmark datasets show significant improvements of MetaShadow over the existing state-of-the-art methods on object-centered shadow detection, removal, and synthesis. MetaShadow excels in image-editing tasks such as object removal, relocation, and insertion, pushing the boundaries of object-centered image editing.

Figures

Figures reproduced from arXiv: 2412.02635 by the authors.

Figure 1
Figure 1. MetaShadow is a versatile three-in-one framework designed for shadow-related tasks, enabling shadow manipulation in various object-centered image editing operations such as: [I] Object Relocation: Our model can detect and remove the shadow of an existing object, then synthesize the shadow in the new location consistent with the original shadow. [II] Remove an object and its shadow: (1) Based on the mask of the unwan… view at source ↗
Figure 2
Figure 2. We construct one training set, i.e., MOS Dataset, and two real-world evaluation sets, i.e., Moving DESOBA Dataset (without ground truths) and Video DESOBA Dataset (with ground truths), to train and evaluate the effectiveness of MetaShadow. as object removal or insertion handle shadows in an implicit manner, forfeiting any controllability on the objects’ effects on the scene, which is crucial in image-editing applica… view at source ↗
Figure 3
Figure 3. The schematic illustration of our MetaShadow framework. In Stage I, the Shadow Analyzer takes the input image with object mask [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Respective limitations of GAN-based and diffusion-based [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Visual comparison for object-centered shadow detection and removal tasks on the DESOBA test set. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Visual comparison for object-centered shadow synthesis on the DESOBA test set [ [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Visual comparison on different random seeds reveals a [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Visualization of the intermediate multi-scale feature maps [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

66 extracted references · 60 canonical work pages

  1. [51]

    Objectdrop: Bootstrap- ping counterfactuals for photorealistic object removal and insertion, 2024

    Daniel Winter, Matan Cohen, Shlomi Fruchter, Yael Pritch, Alex Rav-Acha, and Yedid Hoshen. Objectdrop: Bootstrap- ping counterfactuals for photorealistic object removal and insertion, 2024. 2, 3

    2024

  2. [1]

    eDiff-I: Text-to-image diffusion models with ensemble of expert denoisers

    Yogesh Balaji, Seungjun Nah, Xun Huang, Arash Vahdat, Jiaming Song, Qinsheng Zhang, Karsten Kreis, Miika Ait- tala, Timo Aila, Samuli Laine, Bryan Catanzaro, Tero Kar- ras, and Ming-Yu Liu. eDiff-I: Text-to-image diffusion models with ensemble of expert denoisers. arXiv preprint arXiv:2211.01324, 2022. 8

    arXiv 2022

  3. [2]

    Blender Foundation. Blender. https://www.blender. org, 2023. 3

    work page 2023

  4. [3]

    AnyDoor: Zero-shot object-level image customization

    Xi Chen, Lianghua Huang, Yu Liu, Yujun Shen, Deli Zhao, and Hengshuang Zhao. AnyDoor: Zero-shot object-level image customization. arXiv preprint arXiv:2307.09481, 2023. 8

    arXiv 2023

  5. [4]

    A multi-task mean teacher for semi- supervised shadow detection

    Zhihao Chen, Lei Zhu, Liang Wan, Song Wang, Wei Feng, and Pheng-Ann Heng. A multi-task mean teacher for semi- supervised shadow detection. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5611–5620,

    work page

  6. [5]

    CANet: A context-aware network for shadow removal

    Zipei Chen, Chengjiang Long, Ling Zhang, and Chunxia Xiao. CANet: A context-aware network for shadow removal. In IEEE International Conference on Computer Vision, 2021. 3

    work page 2021

  7. [6]

    Towards ghost- free shadow removal via dual hierarchical aggregation net- work and shadow matting GAN

    Xiaodong Cun, Chi-Man Pun, and Cheng Shi. Towards ghost- free shadow removal via dual hierarchical aggregation net- work and shadow matting GAN. In AAAI Conference on Artificial Intelligence, 2020. 2, 3, 6

    work page 2020

  8. [7]

    ARGAN: Attentive recurrent generative adversarial network for shadow detection and removal

    Bin Ding, Chengjiang Long, Ling Zhang, and Chunxia Xiao. ARGAN: Attentive recurrent generative adversarial network for shadow detection and removal. In IEEE International Conference on Computer Vision, 2019. 3

    work page 2019

Show all 66 references
  1. [8]

    Robust shadow detection by exploring effective shadow con- texts

    Xianyong Fang, Xiaohao He, Linbo Wang, and Jianbing Shen. Robust shadow detection by exploring effective shadow con- texts. In Proceedings of the 29th ACM International Confer- ence on Multimedia, pages 2927–2935, 2021. 2

    2021

  2. [9]

    Magic Eraser

    Google. Magic Eraser. https : / / blog . google / products / photos / magic - eraser - android - ios-google-one/, 2023. 1

    2023

  3. [10]

    ShadowFormer: Global context helps image shadow removal

    Lanqing Guo, Siyu Huang, Ding Liu, Hao Cheng, and Bihan Wen. ShadowFormer: Global context helps image shadow removal. In AAAI Conference on Artificial Intelligence, 2023. 3

    2023

  4. [11]

    ShadowDiffusion: When degradation prior meets diffusion model for shadow removal

    Lanqing Guo, Chong Wang, Wenhan Yang, Siyu Huang, Yufei Wang, Hanspeter Pfister, and Bihan Wen. ShadowDiffusion: When degradation prior meets diffusion model for shadow removal. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. 2, 3, 6

    2023

  5. [12]

    Denoising diffu- sion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffu- sion probabilistic models. In Advances in Neural Information Processing Systems, 2020. 5, 8

    2020

  6. [13]

    Shadow generation for composite image in real-world scenes

    Yan Hong, Li Niu, and Jianfu Zhang. Shadow generation for composite image in real-world scenes. In AAAI Conference on Artificial Intelligence, 2022. 2, 3, 4, 5, 6, 7

    2022

  7. [14]

    Direction-aware spatial context features for shadow detection

    Xiaowei Hu, Lei Zhu, Chi-Wing Fu, Jing Qin, and Pheng-Ann Heng. Direction-aware spatial context features for shadow detection. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7454–7462, 2018. 2

    2018

  8. [15]

    Direction-aware spatial context features for shadow de- tection and removal

    Xiaowei Hu, Chi-Wing Fu, Lei Zhu, Jing Qin, and Pheng-Ann Heng. Direction-aware spatial context features for shadow de- tection and removal. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019. 3

    2019

  9. [16]

    Mask-ShadowGAN: Learning to remove shadows from unpaired data

    Xiaowei Hu, Yitong Jiang, Chi-Wing Fu, and Pheng-Ann Heng. Mask-ShadowGAN: Learning to remove shadows from unpaired data. In IEEE International Conference on Computer Vision, pages 2472–2481, 2019. 3

    2019

  10. [17]

    Revisiting shadow de- tection: A new benchmark dataset for complex world

    Xiaowei Hu, Tianyu Wang, Chi-Wing Fu, Yitong Jiang, Qiong Wang, and Pheng-Ann Heng. Revisiting shadow de- tection: A new benchmark dataset for complex world. IEEE Transactions on Image Processing, 30:1925–1934, 2021. 2, 3

    1925

  11. [18]

    RMLANet: Random multi-level attention network for shadow detection

    Leiping Jie and Hui Zhang. RMLANet: Random multi-level attention network for shadow detection. InIEEE International Conference on Multimedia and Expo, pages 1–6. IEEE, 2022. 2

    2022

  12. [19]

    Yeying Jin, Aashish Sharma, and Robby T. Tan. DC- ShadowNet: Single-image hard and soft shadow removal using unsupervised domain-classifier guided network. In IEEE International Conference on Computer Vision, 2021. 3

    2021

  13. [20]

    Imagic: Text-based real image editing with diffusion models

    Bahjat Kawar, Shiran Zada, Oran Lang, Omer Tov, Hui- wen Chang, Tali Dekel, Inbar Mosseri, and Michal Irani. Imagic: Text-based real image editing with diffusion models. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. 8

    2023

  14. [21]

    Automatic feature learning for robust shadow detection

    Salman Hameed Khan, Mohammed Bennamoun, Ferdous Sohel, and Roberto Togneri. Automatic feature learning for robust shadow detection. In IEEE/CVF Conference on Com- puter Vision and Pattern Recognition, pages 1939–1946, 2014. 2

    1939

  15. [22]

    Shadow removal via shadow image decomposition

    Hieu Le and Dimitris Samaras. Shadow removal via shadow image decomposition. In ICCV, 2019. 2, 3, 5, 6

    2019

  16. [23]

    Yago Vicente, Vu Nguyen, Minh Hoai, and Dimitris Samaras

    Hieu Le, Tom´as F. Yago Vicente, Vu Nguyen, Minh Hoai, and Dimitris Samaras. A+D Net: Training a shadow detector with adversarial shadow attenuation. In European Conference on Computer Vision, pages 662–678, 2018. 2

    2018

  17. [24]

    GLIGEN: Open-set grounded text-to-image genera- tion

    Yuheng Li, Haotian Liu, Qingyang Wu, Fangzhou Mu, Jianwei Yang, Jianfeng Gao, Chunyuan Li, and Yong Jae Lee. GLIGEN: Open-set grounded text-to-image genera- tion. IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. 8

    2023

  18. [25]

    Belongie, Lubomir D

    Tsung-Yi Lin, Michael Maire, Serge J. Belongie, Lubomir D. Bourdev, Ross B. Girshick, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll’a r, and C. Lawrence Zitnick. Microsoft COCO: common objects in context. CoRR, abs/1405.0312,

    arXiv

  19. [26]

    ARShadowGAN: Shadow generative adversarial network for augmented reality in single light scenes

    Daquan Liu, Chengjiang Long, Hongpan Zhang, Hanning Yu, Xinzhi Dong, and Chunxia Xiao. ARShadowGAN: Shadow generative adversarial network for augmented reality in single light scenes. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8139–8148, 2020. 3, 5

    2020

  20. [27]

    A decoupled multi-task network for shadow removal

    Jiawei Liu, Qiang Wang, Huijie Fan, Wentao Li, Liangqiong Qu, and Yandong Tang. A decoupled multi-task network for shadow removal. IEEE Transactions on Multimedia, 2023. 2, 3, 6

    2023

  21. [28]

    Shadow generation for composite image 9 using diffusion model

    Qingyang Liu, Junqi You, Jianting Wang, Xinhao Tao, Bo Zhang, and Li Niu. Shadow generation for composite image 9 using diffusion model. In IEEE/CVF Conference on Com- puter Vision and Pattern Recognition, 2024. 2, 3, 6, 7, 8

    2024

  22. [29]

    From shadow generation to shadow removal

    Zhihao Liu, Hui Yin, Xinyi Wu, Zhenyao Wu, Yang Mi, and Song Wang. From shadow generation to shadow removal. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021. 3

    2021

  23. [30]

    Gustafsson, Zheng Zhao, Jens Sj¨olund, and Thomas B

    Ziwei Luo, Fredrik K. Gustafsson, Zheng Zhao, Jens Sj¨olund, and Thomas B. Sch¨on. Controlling vision-language models for universal image restoration. arXiv preprint arXiv:2310.01018, 2023. 8

    2023 arXiv

  24. [31]

    V-Net: Fully convolutional neural networks for volumetric medical image segmentation

    Fausto Milletari, Nassir Navab, and Seyed-Ahmad Ahmadi. V-Net: Fully convolutional neural networks for volumetric medical image segmentation. In International Conference on 3D Vision, 2016. 5

    2016

  25. [32]

    Yago Vicente, Maozheng Zhao, Minh Hoai, and Dimitris Samaras

    Vu Nguyen, Tom´as F. Yago Vicente, Maozheng Zhao, Minh Hoai, and Dimitris Samaras. Shadow detection with condi- tional generative adversarial networks. In IEEE International Conference on Computer Vision, pages 4510–4518, 2017. 2

    2017

  26. [33]

    Making images real again: A comprehensive survey on deep image composition

    Li Niu, Wenyan Cong, Liu Liu, Yan Hong, Bo Zhang, Jing Liang, and Liqing Zhang. Making images real again: A comprehensive survey on deep image composition. arXiv preprint arXiv:2106.14490, 2021. 6

    2021

  27. [34]

    Liangqiong Qu, Jiandong Tian, Shengfeng He, Yandong Tang, and Rynson W.H. Lau. DeshadowNet: A multi-context em- bedding deep network for shadow removal. In IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 4067–4075, 2017. 3, 5, 6

    2017

  28. [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, 2021. 8

    2021

  29. [36]

    Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J. Liu. Exploring the limits of transfer learning with a unified text-to-text transformer. Journal of Machine Learning Research, 2020. 8

    2020

  30. [37]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, 2022. 5, 8

    2022

  31. [38]

    Mrinmoy Sen, Sai Pradyumna Chermala, Nazrinbanu Nur- mohammad Nagori, Venkat Peddigari, Praful Mathur, B. H. Pawan Prasad, and Moonhwan Jeong. SHARDS: Ef- ficient shadow removal using dual stage network for high- resolution images. In IEEE/CVF Winter Conference on Ap- plicatio...

    2023

  32. [39]

    Shadow opti- mization from structured deep edge detection

    Li Shen, Teck Wee Chua, and Karianto Leman. Shadow opti- mization from structured deep edge detection. In IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 2067–2074, 2015. 2

    2015

  33. [40]

    SSN: Soft shadow network for image compositing

    Yichen Sheng, Jianming Zhang, and Bedrich Benes. SSN: Soft shadow network for image compositing. In IEEE/CVF Conference on Computer Vision and Pattern Recognition ,

  34. [41]

    Cen- giz Oztireli, He Zhang, Zhe Lin, Eli Shechtman, and Bedrich Benes

    Yichen Sheng, Yifan Liu, Jianming Zhang, Wei Yin, A. Cen- giz Oztireli, He Zhang, Zhe Lin, Eli Shechtman, and Bedrich Benes. Controllable shadow generation using pixel height maps. In European Conference on Computer Vision, 2022. 2

    2022

  35. [42]

    PixHt-Lab: Pixel height based light effect generation for image compositing

    Yichen Sheng, Jianming Zhang, Julien Philip, Yannick Hold- Geoffroy, Xin Sun, He Zhang, Lu Ling, and Bedrich Benes. PixHt-Lab: Pixel height based light effect generation for image compositing. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. 2, 3

    2023

  36. [43]

    Objectstitch: Object compositing with diffusion model

    Yizhi Song, Zhifei Zhang, Zhe Lin, Scott Cohen, Brian Price, Jianming Zhang, Soo Ye Kim, and Daniel Aliaga. Objectstitch: Object compositing with diffusion model. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. 3, 8

    2023

  37. [44]

    Yago Vicente, Le Hou, Chen-Ping Yu, Minh Hoai, and Dimitris Samaras

    Tom´as F. Yago Vicente, Le Hou, Chen-Ping Yu, Minh Hoai, and Dimitris Samaras. Large-scale training of shadow detec- tors with noisily-annotated shadow examples. In European Conference on Computer Vision, pages 816–832, 2016. 2, 3

    2016

  38. [45]

    Style-guided shadow removal

    Jin Wan, Hui Yin, Zhenyao Wu, Xinyi Wu, Yanting Liu, and Song Wang. Style-guided shadow removal. In European Conference on Computer Vision, 2022. 2, 3

    2022

  39. [46]

    Stacked conditional generative adversarial networks for jointly learning shadow detection and shadow removal

    Jifeng Wang, Xiang Li, and Jian Yang. Stacked conditional generative adversarial networks for jointly learning shadow detection and shadow removal. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1788–1797,

  40. [47]

    Instance shadow detection

    Tianyu Wang, Xiaowei Hu, Qiong Wang, Pheng-Ann Heng, and Chi-Wing Fu. Instance shadow detection. In IEEE/CVF Conference on Computer Vision and Pattern Recognition ,

  41. [48]

    Instance shadow detection

    Tianyu Wang∗, Xiaowei Hu∗, Qiong Wang, Pheng-Ann Heng, and Chi-Wing Fu. Instance shadow detection. In IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 1880–1889, 2020. * Joint first authors. 3

    2020

  42. [49]

    Single-stage instance shadow detection with bidirec- tional relation learning

    Tianyu Wang, Xiaowei Hu, Chi-Wing Fu, and Pheng-Ann Heng. Single-stage instance shadow detection with bidirec- tional relation learning. In IEEE/CVF Conference on Com- puter Vision and Pattern Recognition, pages 1–11, 2021. 2, 6

    2021

  43. [50]

    Instance shadow detection with a single-stage detector

    Tianyu Wang, Xiaowei Hu, Pheng-Ann Heng, and Chi-Wing Fu. Instance shadow detection with a single-stage detector. IEEE Transactions on Pattern Analysis and Machine Intelli- gence, pages 1–14, 2022. 2, 6

    2022

  44. [52]

    Light-weight shadow detection via GCN-based annotation strategy and knowledge distillation

    Wen Wu, Kai Zhou, Xiao-Diao Chen, and Jun-Hai Yong. Light-weight shadow detection via GCN-based annotation strategy and knowledge distillation. Computer Vision and Image Understanding, 216:103341, 2022. 2

    2022

  45. [53]

    Paint by ex- ample: Exemplar-based image editing with diffusion models

    Binxin Yang, Shuyang Gu, Bo Zhang, Ting Zhang, Xuejin Chen, Xiaoyan Sun, Dong Chen, and Fang Wen. Paint by ex- ample: Exemplar-based image editing with diffusion models. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. 8

    2023

  46. [54]

    SILT: Shadow-aware iterative label tuning for learning to detect shadows from noisy labels

    Han Yang, Tianyu Wang, Xiaowei Hu, and Chi-Wing Fu. SILT: Shadow-aware iterative label tuning for learning to detect shadows from noisy labels. In IEEE International Conference on Computer Vision, 2023. 1, 2 10

    2023

  47. [55]

    LRA&LDRA: Rethinking residual predictions for efficient shadow detection and removal

    Mehmet Kerim Y¨ucel, Valia Dimaridou, Bruno Manganelli, Mete Ozay, Anastasios Drosou, and Albert Sa `a-Garriga. LRA&LDRA: Rethinking residual predictions for efficient shadow detection and removal. In IEEE/CVF Winter Confer- ence on Applications of Computer Vision, 2023. 3

    2023

  48. [56]

    Edward Zhang, Ricardo Martin-Brualla, Janne Kontkanen, and Brian L. Curless. No shadow left behind: Removing objects and their shadows using approximate lighting and geometry. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021. 2, 3

    2021

  49. [57]

    Adding conditional control to text-to-image diffusion models

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

    2023

  50. [58]

    Shadow- GAN: Shadow synthesis for virtual objects with conditional adversarial networks

    Shuyang Zhang, Runze Liang, and Miao Wang. Shadow- GAN: Shadow synthesis for virtual objects with conditional adversarial networks. CVM, 2019. 3

    2019

  51. [59]

    CM-GAN: Image inpainting with cascaded modulation GAN and object-aware training

    Haitian Zheng, Zhe Lin, Jingwan Lu, Scott Cohen, Eli Shecht- man, Connelly Barnes, Jianming Zhang, Ning Xu, Sohrab Amirghodsi, and Jiebo Luo. CM-GAN: Image inpainting with cascaded modulation GAN and object-aware training. In European Conference on Computer Vision, 2022. 4, 5

    2022

  52. [60]

    Quanlong Zheng, Xiaotian Qiao, Ying Cao, and Rynson W.H. Lau. Distraction-aware shadow detection. In IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 5167–5176, 2019. 2

    2019

  53. [61]

    Shadow detection via multi-scale feature fusion and unsupervised domain adaptation

    Kai Zhou, Wen Wu, Yan-Li Shao, Jing-Long Fang, Xing-Qi Wang, and Dan Wei. Shadow detection via multi-scale feature fusion and unsupervised domain adaptation. Journal of Visual Communication and Image Representation, 88:103596, 2022

    2022

  54. [62]

    Bidirectional feature pyramid network with recurrent attention residual modules for shadow detection

    Lei Zhu, Zijun Deng, Xiaowei Hu, Chi-Wing Fu, Xuemiao Xu, Jing Qin, and Pheng-Ann Heng. Bidirectional feature pyramid network with recurrent attention residual modules for shadow detection. In European Conference on Computer Vision, pages 121–136, 2018

    2018

  55. [63]

    Lei Zhu, Ke Xu, Zhanghan Ke, and Rynson W.H. Lau. Miti- gating intensity bias in shadow detection via feature decom- position and reweighting. In IEEE International Conference on Computer Vision, pages 4702–4711, 2021. 1

    2021

  56. [64]

    Single image shadow detection via complementary mechanism

    Yurui Zhu, Xueyang Fu, Chengzhi Cao, Xi Wang, Qibin Sun, and Zheng-Jun Zha. Single image shadow detection via complementary mechanism. In Proceedings of the 30th ACM International Conference on Multimedia, pages 6717–6726,

  57. [65]

    Bijective mapping network for shadow removal

    Yurui Zhu, Jie Huang, Xueyang Fu, Feng Zhao, Qibin Sun, and Zheng-Jun Zha. Bijective mapping network for shadow removal. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022. 2, 3

    2022

  58. [66]

    Efficient model-driven network for shadow removal

    Yurui Zhu, Zeyu Xiao, Yanchi Fang, Xueyang Fu, Zhiwei Xiong, and Zheng-Jun Zha. Efficient model-driven network for shadow removal. In AAAI Conference on Artificial Intelli- gence, 2022. 2, 3 11

    2022

Pith tools

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