Comment by alexcnwy

First of all thank you for all your hard work.

Major question:

Why isn’t Tesseract using neural networks. I know it just introduced LSTM based models but they suck.

Why is GCP vision text recognition API so much better than open source alternatives?!