Hello,
First of all let me start by saying that I greatly appreciate the amount of effort put into tesseract, how accurate it has become and the fact that is is free :-)
I've read various topics about GPU support, and I understand that at some point OpenCL was planned, but didn't really go through?
Since the newer versions use LSTM, I have to ask, are there any plans to offer CUDA support for training and/or evaluating (batch) documents?
From my limited understanding of LSTM I would have assumed that might make sense, although I also understand that the effort behind doing that would be huge.
I am at a point where tesseract has gotten me good results with a variety of languages, but I can't use it in real-time applications because it seems to
take on average 15 seconds per query, and I have been asked to reduce this to under a second.
I apologise if this has been discussed before, I did try searching the forum first.
Regards,
Alex