I don't think so But anyways just a side note one way to help reduce these errors is if you pass in both the original image and the OCR'd text to the models th

I don't think so But anyways just a side note one way to help reduce these errors is if you pass in both the original image and the OCR'd text to the models that make the decisions