@bdenckla By WLC, do you mean UXLC or WLC 4.20 (the last fork available on https://t.co/JuULcKDLES)? Is WLC in unicode available anywhere even? The Groves Center site hasn't been updated in years
https://t.co/W1bEGAIq1E
The most defining factor of this project is what it is not. It is *not* OCR…sort of. Optical Character Recognition is about determining which letters are on an image and where they are. This digital edition is about encoding the semantics of the layout 🧵
Throughout the week, I hope to explain what sets this project apart from other related (and complementary) projects. But for now users can view individual pages, hover over text and see it in a mss image, and double click to zoom.
https://t.co/3KZQkWtzvd
Though vision capable AI models were used, meaning there was some level of OCR ability on the part of the models, this is not an OCR project. Models were given the text beforehand and told the lay it out. They could read the text, but not creating zones.
Though vision capable AI models were used, meaning there was some level of OCR ability on the part of the models, this is not an OCR project.
Models were given the text beforehand and told the lay it out. They could read the text, but not creating zones.