I’m facing a problem where I got bunch of form scans. The problem is that by applying OCR tool it has dificulties with reading it because of overlapping text. I would like to delete the form schema and leave only input provided by people.
What do you guys think, what would be the best solution for that problem?
I don’t know if the user input is handwritten or not but generally check if you find some useful paper in DI@KDD2021 or in the previous edition of this workshop.