Remove shadows in images using weak supervision with UnShadowNet
Remove shadows in images using weak supervision with UnShadowNet
UnShadowNet: Illumination Critic Guided Contrastive Learning For Shadow Removal
arXiv paper abstract https://arxiv.org/abs/2203.15441v1
arXiv PDF paper https://arxiv.org/pdf/2203.15441v1.pdf
Shadows are frequently encountered natural phenomena that significantly hinder the performance of computer vision perception systems in practical settings, e.g., autonomous driving.
… eliminate shadow … requires pairs of aligned shadowed and non-shadowed images which are difficult to obtain.
… introduce a novel weakly supervised shadow removal framework UnShadowNet trained using contrastive learning.
… comprises of a DeShadower network responsible for removal of the extracted shadow under the guidance of an Illumination network which is trained adversarially by the illumination critic and a Refinement network to further remove artifacts.
… UnShadowNet can … be … extended to a fully-supervised setup to exploit the ground-truth when available.
UnShadowNet outperforms … state-of-the-art … on three publicly available shadow datasets … in both the weakly and fully supervised setups.
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
