TASHR

Text-Aware Single Image Specular Highlight Removal

Dataset

We collect three high-quality datasets with fine-grained annotations, you can download through the following link:

• SD1(Size=10.4G) Google Drive, Baidu Pan(password:8rqp)
• SD2(Size=12.3G) Google Drive, Baidu Pan(password:q3mz)
• RD Baidu Pan(password:uamp)

Citation

If any part of our paper and repository is helpful to your work, please generously cite with:

@inproceedings{hou2021tashr,
  author = {Shiyu Hou, Chaoqun Wang, Weize Quan, Jingen Jiang, and Dong-Ming Yan},
  title = {Text-Aware Single Image Specular Highlight Removal},
  booktitle = {The Proceedings of The Chinese Conference on Pattern Recognition and Computer Vision (PRCV)},
  year = {2021}
}