Implement ARU-Net instead of PRHLT [Suggestion]
Opened this issue · 1 comments
Deleted user commented
.
lquirosd commented
Thanks for the suggestion. My understanding is ARU-Net main goal is to detect the lines/baselines on the documents. Our main goal is slightly different, we want to extract not only the lines/baselines but the structure of the document in a integrated way and this repository is a window to the work we are doing in that direction.