Hi everyone, I'm working on a document understanding project and would appreciate some advice on the right technical direction. The input will be scanned pages or images from academic books. I don't know in advance what kind of figures they'll contain—they could be biology diagrams, anatomy illustra
