Adding more margin to the output squares?
Hi,
I really appreciate the incredible work you've done! I've tested it with my documents, and it's working beautifully🌄
I do have one question: Is there a way to slightly increase the width and height of the output boxes by a few pixels? While most of the output boxes fit well with the rest of the text columns, some boxes—particularly those containing formulas and headers—seem to be just a bit too small by a few pixels. Are there any parameters I could adjust to enlarge specific types of output boxes?
Thanks in advance for your help.
Kind regards,
Hi,
Thank you for your kind words and your interest in our project!
Unfortunately we do not offer some custom parameters to change the visualization but you can look inside src/visualization/save_output_to_pdf.py. Inside of this file, there is a method called add_prediction_annotation. You can try to tweak the parameters there according to your needs.
After changing the parameters, you should build the docker image from source like:
make start
Best
Hi.
I've been looking into OCR results compared to the Upstage's documentparse, I think this one is more accurate in overall. (Although theirs runs 100x faster on their machine)
Although I can't provide benchmark metrics, I just wanted to let you know.
Best regards.