mlsanityData to 7 Oct 2026

Single-Image Shadow Removal Using Deep Learning: A Comprehensive Survey

arXiv 2407.08865 ↗First posted 11 Jul 2024cs.CVcited by 7

Not on the map. Why?

Abstract (arXiv)

Shadow removal aims at restoring the image content within shadow regions, pursuing a uniform distribution of illumination that is consistent between shadow and non-shadow regions. {Comparing to other image restoration tasks, there are two unique challenges in shadow removal:} 1) The patterns of shadows are arbitrary, varied, and often have highly complex trace structures, making ``trace-less'' image recovery difficult. 2) The degradation caused by shadows is spatially non-uniform, resulting in inconsistencies in illumination and color between shadow and non-shadow areas. Recent developments in this field are primarily driven by deep learning-based solutions, employing a variety of learning strategies, network architectures, loss functions, and training data. Nevertheless, a thorough and insightful review of deep learning-based shadow removal techniques is still lacking. In this paper, we are the first to provide a comprehensive survey to cover various aspects ranging from technical details to applications. We highlight the major advancements in deep learning-based single-image shadow removal methods, thoroughly review previous research across various categories, and provide insights into the historical progression of these developments. Additionally, we summarize performance comparisons both quantitatively and qualitatively. Beyond the technical aspects of shadow removal methods, we also explore potential future directions for this field.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 52 of its 120 references that are in the map. First posted on arXiv 11 Jul 2024.

Builds on

52 of 120 references.

References · 52

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  4. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  5. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  6. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  7. 2023Segment Anythingcited by 8,158
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  9. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  10. 2022Generative Adversarial Networkscited by 5,603
  11. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,290
  12. 2021SwinIR: Image Restoration Using Swin Transformercited by 1,869
38 more
  1. 2021Image Super-Resolution via Iterative Refinementcited by 1,506
  2. 2022RePaint: Inpainting using Denoising Diffusion Probabilistic Modelscited by 1,419
  3. 2022Denoising Diffusion Restoration Modelscited by 951
  4. 2020Zero-Reference Deep Curve Estimation for Low-Light Image Enhancementcited by 541
  5. 2023Exploiting Diffusion Prior for Real-World Image Super-Resolutioncited by 486
  6. 2017"Zero-Shot" Super-Resolution using Deep Internal Learningcited by 285
  7. 2020Deep Unfolding Network for Image Super-Resolutioncited by 253
  8. 2017Stacked Conditional Generative Adversarial Networks for Jointly Learning Shadow Detection and Shadow Removalcited by 160
  9. 2021Low-Light Image and Video Enhancement Using Deep Learning: A Surveycited by 148
  10. 2022Deep Generalized Unfolding Networks for Image Restorationcited by 119
  11. 2019Towards Ghost-free Shadow Removal via Dual Hierarchical Aggregation Network and Shadow Matting GANcited by 102
  12. 2022ShadowDiffusion: When Degradation Prior Meets Diffusion Model for Shadow Removalcited by 102
  13. 2022Shadows can be Dangerous: Stealthy and Effective Physical-world Adversarial Attack by Natural Phenomenoncited by 101
  14. 2019Shadow Removal via Shadow Image Decompositioncited by 96
  15. 2019Mask-ShadowGAN: Learning to Remove Shadows from Unpaired Datacited by 91
  16. 2021Auto-Exposure Fusion for Single-Image Shadow Removalcited by 78
  17. 2022DC-ShadowNet: Single-Image Hard and Soft Shadow Removal Using Unsupervised Domain-Classifier Guided Networkcited by 68
  18. 2020From Shadow Segmentation to Shadow Removalcited by 59
  19. 2018Direction-aware Spatial Context Features for Shadow Detection and Removalcited by 58
  20. 2021From Shadow Generation to Shadow Removalcited by 54
  21. 2019ARGAN: Attentive Recurrent Generative Adversarial Network for Shadow Detection and Removalcited by 51
  22. 2021CANet: A Context-Aware Network for Shadow Removalcited by 47
  23. 2020Shadow Removal by a Lightness-Guided Network with Training on Unpaired Datacited by 38
  24. 2022DeS3: Adaptive Attention-driven Self and Soft Shadow Removal using ViT Similaritycited by 37
  25. 2019RIS-GAN: Explore Residual and Illumination with Generative Adversarial Networks for Shadow Removalcited by 35
  26. 2021Shadow Generation for Composite Image in Real-world Scenescited by 35
  27. 2017Improving Shadow Suppression for Illumination Robust Face Recognitioncited by 29
  28. 2023ExposureDiffusion: Learning to Expose for Low-light Image Enhancementcited by 29
  29. 2023Latent Feature-Guided Diffusion Models for Shadow Removalcited by 28
  30. 2021Learning from Synthetic Shadows for Shadow Detection and Removalcited by 26
  31. 2023Leveraging Inpainting for Single-Image Shadow Removalcited by 24
  32. 2024Recasting Regional Lighting for Shadow Removalcited by 23
  33. 2018Conditional GANs for Multi-Illuminant Color Constancy: Revolution or Yet Another Approach?cited by 17
  34. 2021Unsupervised Portrait Shadow Removal via Generative Priorscited by 16
  35. 2024Progressive Divide-and-Conquer via Subsampling Decomposition for Accelerated MRIcited by 13
  36. 2016Interactive Removal and Ground Truth for Difficult Shadow Scenescited by 11
  37. 2024Benchmarking Adversarial Robustness of Image Shadow Removal with Shadow-adaptive Attackscited by 10
  38. 2023Shadow Generation with Decomposed Mask Prediction and Attentive Shadow Fillingcited by 8

50 most cited of 52 All 52 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 2; 2026: 3.

7 papers.

  1. 2024Unveiling Deep Shadows: A Survey and Benchmark on Image and Video Shadow Detection, Removal, and Generation in the Deep Learning Eracited by 6
  2. 2025DenseSR: Image Shadow Removal as Dense Predictioncited by 5
  3. 2024SoftShadow: Leveraging Soft Masks for Penumbra-Aware Shadow Removalcited by 4
  4. 2025FASR-Net: Unsupervised Shadow Removal Leveraging Inherent Frequency Priorscited by 1
  5. 2026UnSCAR: Universal, Scalable, Controllable, and Adaptable Image Restorationcited by 0
  6. 2026FreeShadow: Training-Free Shadow Removal via Illumination Transfer and Selective Content Preservation in Diffusion Modelscited by 0
  7. 2026Consistency as Regularization for Unsupervised Shadow Removalcited by 0