Skip to content

Latest commit

 

History

History
24 lines (15 loc) · 1.06 KB

README.md

File metadata and controls

24 lines (15 loc) · 1.06 KB

HTR-using-CRNN

A school Project for Optical Character:wORDS Recogniton more oriented to handwritting recongnition done using Convolutional Recurrent Neural Networks(CRNN) using The IAM dataset

I trained the model thanks to GoogleColab reached an accuracy of 99.99%

Model Accuracy

The Model was able to spot some words like this :

The Model also failed to recongnize some other tedious words like this: