Shadow feature refinement network improves shadow removal accuracy
Shadow Feature Refinement Network: Progressive Feature Refinement based on Knowledge Distillation for Effective Shadow Removal
Computer Vision and Pattern Recognition
Summary
Removing shadows from photos is hard because shadows come in many sizes and colors. The authors introduced a new computer program called Shadow Feature Refinement Network (SFR-Net) that learns to remove shadows more effectively by focusing on refining features and using guidance from a teacher model. They also developed a way to adjust the colors after removing shadows so images look natural. When tested on standard shadow image datasets, their method worked well and produced clear, shadow-free pictures.
What this means in practice
- •For photo editing software developers: Integrate SFR-Net to provide more accurate shadow removal tools that preserve natural colors and improve image quality automatically.$Commercial implications: Enables photo editing products to offer advanced shadow removal features that improve user experience and output quality.
- •For autonomous vehicle perception teams: Enhance vehicle vision systems by removing shadows from camera images, improving object detection under variable lighting conditions.
Authors
Donghyun Han, Byoung-Dai Lee
Abstract
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.