OCR is considered a solved problem in general but not in entirety
A key component of it, Handwriting recognition is still a challenging problem.
Handwriting Text Recognition(HTR) is the task of recognizing handwritten human text
It involves using both Computer Vision and NLP
Every person has a different style of handwriting, thus solving the task of HTR is much more difficult than OCR
In this article I cover the progress of various techniques involved in solving HTR and all the SOTA models
In addition, I have discussed the way to train your own HTR model on your own dataset
Happy to discuss more if you have any queries about handwritten text recognition