Compare and Evaluate Pre-Trained Models on Text Extraction for Passport Reader

[abccastro]

Finalize the selection of the pre-trained model for extracting text from US passports by comparing and evaluating the accuracy of results from both PaddleOCR and Pytesseract

[abccastro]

As discussed with the Professor (or the client), we will drop this feature as the results from the pre-trained models are not accurate enough to properly extract the texts from the passports.