Shadow Detection and Removal

Shadow removal 


Shadows are ubiquitous in images of natural scenes. On one hand, shadows provide useful cues about the scene including object shapes, light sources and illumination conditions, camera parameters and geo-location, and scene geometry. On the other hand, the presence of shadows in images creates difficulties for many computer vision tasks from image segmentation to object detection and tracking. In all cases, being able to automatically detect shadows, and subsequently remove them or reason about their shapes and sizes would usually be beneficial. Moreover, shadow-free images are of great interest for image editing, computational photography, and augmented reality, and the first crucial step is shadow detection.

The aim of this project is to detect and remove shadows in still images. Our project has so far produced: 1) an algorithm to optimize the relative importance of different feature cues for shadow detection; 2) a lazy labeling tool for quickly annotate shadow images; 3) a large-scale shadow detaset; and 4) a deep-learning method for learning from noisily-annotated shadow images.


  • Tomas Yago Vicente, PhD student

  • Vu Nguyen, PhD student

  • Maozheng Zhao, PhD student

  • Hieu Le, PhD student


  • A+D Net: Training a Shadow Detector with Adversarial Shadow Attenuation.
    Le, H., Vicente, T.F.Y., Nguyen, V., Hoai, M. & Samaras, D. (2018)
    Proceedings of European Conference on Computer Vision (ECCV).
    Paper BibTex.

  • Leave-one-out Kernel Optimization for Shadow Detection and Removal.
    Vicente, T.F.Y., Hoai, M., & Samaras, D. (2018)
    IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI) 40(3), 682–695.
    Paper BibTex.

  • Shadow Detection with Conditional Generative Adversarial Networks.
    Nguyen, V., Vicente, T., Zhao, M., Hoai, M. & Samaras, D. (2017)
    Proceedings of International Conference on Computer Vision (ICCV).
    Paper Project BibTex.

  • Large-scale Training of Shadow Detectors with Noisily-Annotated Shadow Examples.
    Vicente, T.F.Y., Hou, L., Yu, C.-P., Hoai, M. & Samaras, D. (2016)
    Proceedings of European Conference on Computer Vision (ECCV).
    Paper BibTex.

  • Noisy Label Recovery for Shadow Detection in Unfamiliar Domains.
    Vicente, T.F.Y., Hoai, M., & Samaras, D. (2016)
    Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
    Paper BibTex.

  • Leave-One-Out Kernel Optimization for Shadow Detection.
    Vicente, T.F.Y., Hoai, M., & Samaras, D. (2015)
    Proceedings of International Conference on Computer Vision (ICCV).
    Paper BibTex.


  • SBU Shadow dataset (286MB). This dataset consists of around 5000 images containing shadows from a wide variety of scenes and photo types. Annotations in the form of shadow binary masks are provided along with the actual images. The shadow label annotations for the 4000 images in the training set are the result of applying our proposed label recovery method to reduce label noise. Whereas, the testing images were carefully annotated manually to produce precise shadow masks.

Copyright notice

The documents contained in these directories are included by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a non-commercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.