Currently we are using trocr-base-printed. Identify 3 other models and evaluate our OCR pipeline using these models. The model can be changed in the ocr/services/image_ocr.py file.
Metrics : Feel free to use the functions in the metrics class to do this. services/metrics_analysis.py
Data: Use at least 1 Fillable, 1 Non Fillable, 1 Lab Form, 1 Case Report form for each model
Location of all Forms, Ground Truth and Segmented Files
Currently we are using trocr-base-printed. Identify 3 other models and evaluate our OCR pipeline using these models. The model can be changed in the ocr/services/image_ocr.py file.
Metrics : Feel free to use the functions in the metrics class to do this. services/metrics_analysis.py Data: Use at least 1 Fillable, 1 Non Fillable, 1 Lab Form, 1 Case Report form for each model
Location of all Forms, Ground Truth and Segmented Files
https://drive.google.com/drive/folders/1vG-XwFiOoRunx9eXCmd41WpAI5NCizeM