Bounding Box for pdf using Flash 2.0

Gemini API
1

I tried two ways to create bounding box for pdf (which is some kind of form):

  1. directly sending the pdf as bytes
  2. converting each page to image and then sending

2nd option give result better than the 1st one, I am not sure about the reason. Is there any better way to do it as 2nd option requires some extra processing?

3

Hi @gauraww,

Which model you are using? Try to use the latest model with updated one like 2.5 pro. It also depends on prompting. Do some prompt iteration , refine your prompt and see if you are getting better result. If none of this working, it might be model response issue.

4

Currently using Flash 2.0. Tried various prompt iteration, seems to be consistent

5

Try with 2.5, otherwise it might be a model quality issue, for that we need to wait for next version release with improved capability.