> > Fine Tune will work if all you want to change is a font, with the same > unicharset. This works well for Latin script based languages but not > complex scripts. >
So you mean it is impossible for fine-turning to learn a new word whose characters are new to the LSTM net? If it's true, I am very disappointed but I have no idea what causes this limitation because LSTM net only learn the mapping between image and text. Btw, if it's true there is no chance for empty LSTM net to learn the first word? (i.e. learning from scratch) -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/tesseract-ocr. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/b1ed3ae1-1bd1-491b-9d68-63e554306c77%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
