I was trying to understand how Handwritten Text Recognition works but here I am. I did a lot of research but still, I couldn't exactly understand how will HTR architectures work even with different chars than ASCII. For example, I saw this GitHub project new implementation of HTR but I am not very sure that if I add for example "Turkish chars" to my charset and train model with Turkish data it will work. If it will work, there is no problem. But if it won't, what should I do to create an HTR system for recognizing Turkish letters? Thanks.

  • $\begingroup$ Welcome to DataScienceSE. I'm not an expert in HTR but normally there's no reason for the model to care about the language or char set. But the training data matters of course: you would need to provide the model with a large number of samples (handwritten text as an image + result as text). $\endgroup$
    – Erwan
    Nov 20, 2021 at 16:56


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.