The specific network structure is particular to a given traineddata file. I have no idea how the specification gets turned into the entire network's architecture in tesseract 4, but you can get a particular file's specification with combine_tessdata:
$ ./combine_tessdata.exe -d tess4traineddata/eng.traineddata Version string:4.00.00alpha:eng:synth20170629:[1,36,0,1Ct3,3,16Mp3,3Lfys64Lfx96Lrx96Lfx512O1c1] 17:lstm:size=11689099, offset=192 18:lstm-punc-dawg:size=4322, offset=11689291 19:lstm-word-dawg:size=3694794, offset=11693613 20:lstm-number-dawg:size=4738, offset=15388407 21:lstm-unicharset:size=6360, offset=15393145 22:lstm-recoder:size=1012, offset=15399505 23:version:size=80, offset=15400517 The specification there is " [1,36,0,1Ct3,3,16Mp3,3Lfys64Lfx96Lrx96Lfx512O1c1]". No idea what that means, but the wiki has more to say about it here: https://github.com/tesseract-ocr/tesseract/wiki/TrainingTesseract-4.00#training-just-a-few-layers On Friday, January 12, 2018 at 1:46:38 AM UTC-5, sujith vemisetty wrote: > > I have tried a lot to find the network architecture of LSTMs used in > Tesseract 4.00Alpha, but I wasn't able to find any. I can only find how to > train the new neural network implementation. I would like to understand the > architecture first. Can anyone point me to any documentation which details > the layers of LSTM network, if there is any available? Thanks in advance. > -- 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 tesseract-ocr+unsubscr...@googlegroups.com. To post to this group, send email to tesseract-ocr@googlegroups.com. 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/cf5c9a71-a09e-49cd-8672-e3c892bf2903%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
