auto-grader icon indicating copy to clipboard operation
auto-grader copied to clipboard

Published 20 hours ago • verified publisher icon IEEE-VIT

Improve the character recognition model/ pipeline

Open iamyajat opened this issue 3 years ago • 2 comments

Right now, the character recognition model works very well on the training data but fails to perform as good on the testing data. Improve the accuracy of the whole OCR pipeline (extraction and preprocessing of the boxes and the classification model itself)

iamyajat avatar Sep 30 '21 20:09 iamyajat

Where can I find the dataset?

CaptAlpha avatar Oct 02 '21 17:10 CaptAlpha

Kaggle A-Z dataset

iamyajat avatar Oct 02 '21 18:10 iamyajat