This repository provides the dewarping results for Marior (ACMMM 2022) on DocUNet dataset, which can be downloaded here.
The further cropped images are obtained by runing get_further_crop_im.py.
@inproceedings{JiaxinACMMM2022,
Author = {Jiaxin Zhang, Canjie Luo, Lianwen Jin*, Fengjun Guo, Kai Ding*},
Booktitle = {Proceedings of the 30th ACM International Conference on Multimedia},
Title = {Marior: Margin Removal and Iterative Content Rectification for Document Dewarping in the Wild},
Year = {2022}}