Thanks, but not exactly what I was asking about. If I take the previously trained dataset (dataset A), produced from the text A, and use it to recognise a new text (text B) *and* merge the corrected result (dataset B) with the dataset A, getting "dataset A + dataset B" -- would it benefit an accuracy of the subsequent recognising sessions? Even the accuracy of the recognition of the text B with the "dataset A + dataset B"?
Right now I'm getting about 100% accuracy on the text A with the dataset A, but somehow can't visibly improve the accuracy for the text B, even with the "dataset A + dataset B". Ray Smith wrote: > Running the same data through the training system multiple times does > not change accuracy in tesseract. It does not use a back-propagation > training process at this time. ... --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/tesseract-ocr?hl=en -~----------~----~----~----~------~----~------~--~---
