mlsanityData to 14 Oct 2026

Shadow Feature Refinement Network: Progressive Feature Refinement based on Knowledge Distillation for Effective Shadow Removal

arXiv 2610.05325 ↗First posted 4 Oct 2026cs.CVcited by 0

On the map: Knowledge distillation → Detection and dense prediction distillation

Abstract (arXiv)

In the field of deep learning, has seen significant advancements; however, shadow removal remains a persistent challenge owing to the variable sizes and colors of shadows influenced by lighting conditions. This study proposes a novel shadow feature refinement network (SFR-Net), which leverages supervised learning, feature refinement loss, and knowledge distillation to enhance shadow removal performance. A dedicated post-processing algorithm is further introduced to restore natural color consistency in the generated shadow-free images. We evaluated our method on two public datasets: the adjusted image shadow triplet dataset (ISTD+) and the shadow removal dataset (SRD), which demonstrate strong generalization capabilities under diverse conditions. On ISTD+, our model achieved a root mean square error (RMSE) of 3.4627 and structural similarity index measure (SSIM) of 0.9382 across the entire image. On SRD, it recorded an RMSE of 4.3781 and an SSIM of 0.9341. These comprehensive results show that our approach performs competitively across both shadow and non-shadow regions while setting a promising direction for robust and perceptually natural shadow removal. Code is available at https://github.com/DongHyun99/SFRNet.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Detection and dense prediction distillation, one of 12 directions in Knowledge distillation. Builds on 25 of its 48 references that are in the map. First posted on arXiv 4 Oct 2026.

Builds on

25 of 48 references.

References · 25

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  3. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,859
  4. 2016Perceptual Losses for Real-Time Style Transfer and Super-Resolutioncited by 2,694
  5. 2013Rich feature hierarchies for accurate object detection and semantic segmentationcited by 2,162
  6. 2017High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANscited by 1,179
  7. 2021Resolution-robust Large Mask Inpainting with Fourier Convolutionscited by 738
  8. 2017Generative Adversarial Networks: An Overviewcited by 590
  9. 2016One-Shot Video Object Segmentationcited by 235
  10. 2017Stacked Conditional Generative Adversarial Networks for Jointly Learning Shadow Detection and Shadow Removalcited by 161
  11. 2019Towards Ghost-free Shadow Removal via Dual Hierarchical Aggregation Network and Shadow Matting GANcited by 103
  12. 2022ShadowDiffusion: When Degradation Prior Meets Diffusion Model for Shadow Removalcited by 103
13 more
  1. 2019Shadow Removal via Shadow Image Decompositioncited by 98
  2. 2019Mask-ShadowGAN: Learning to Remove Shadows from Unpaired Datacited by 92
  3. 2021Auto-Exposure Fusion for Single-Image Shadow Removalcited by 79
  4. 2020Improved Techniques for Training Single-Image GANscited by 78
  5. 2022DC-ShadowNet: Single-Image Hard and Soft Shadow Removal Using Unsupervised Domain-Classifier Guided Networkcited by 69
  6. 2021From Shadow Generation to Shadow Removalcited by 55
  7. 2020Physics-based Shadow Image Decomposition for Shadow Removalcited by 39
  8. 2022DeS3: Adaptive Attention-driven Self and Soft Shadow Removal using ViT Similaritycited by 38
  9. 2023Latent Feature-Guided Diffusion Models for Shadow Removalcited by 29
  10. 2021Learning from Synthetic Shadows for Shadow Detection and Removalcited by 27
  11. 2023Leveraging Inpainting for Single-Image Shadow Removalcited by 25
  12. 2022CRFormer: A Cross-Region Transformer for Shadow Removalcited by 8
  13. 2016Appearance Harmonization for Single Image Shadow Removalcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024Detail-Preserving Latent Diffusion for Stable Shadow Removal
  2. 2026After a Decade: Bringing Shadow Removal into the Real World with Agentic Training Data
  3. 2026FreeShadow: Training-Free Shadow Removal via Illumination Transfer and Selective Content Preservation in Diffusion Models
  4. 2022Rich Feature Distillation with Feature Affinity Module for Efficient Image Dehazing
  5. 2025DeshadowMamba: Deshadowing as 1D Sequential Similarity
  6. 2026OmniLight: One Model to Rule All Lighting Conditions
  7. 2026ODDR: One-Step Deshadow Diffusion via Reward Guidance
  8. 2023Deshadow-Anything: When Segment Anything Model Meets Zero-shot shadow removal