Understanding annotation structure in historical datasets

[shubh umre]

Hi,

While exploring the project idea around recovering historical annotations, I was wondering — are there any existing examples of partially reconstructed datasets or intermediate formats that could be used as a reference?

I’m trying to get a clearer picture of how the raw annotated images are typically transformed into structured data in this context.

Thanks!