Hi Team,
In our case we are receiving PDF which are works well with DetectMode PHOTO since it has Diacom/XRAY images in them but same PDF do not work well with CURVED_TEXT Mode .
The overall accuracy of text decrease which I can understand because of different purpose served by DETECT mode.
Is it possible to get a confidence score on recognized page result. This will help us to switch the DETECT Mode are runtime based on that score