Hello everyone ,
I 'm working on an OCR project I want to extract all the informations in arabic from id card i found easyocr (lang=‘ar’) but the detction was not good at all and the output is very messy I 'm thinking about finetuning easyocr if anyone can help me ? I used also gemini pro vision API and that gives perfect result but how i want a local mutimodal that do the same thing or similar i need help thanks in advance