Closed richpomfret closed 6 years ago
A computer vision project for FreeCEN is detailed on FreeCEN2 Issue 393
Open question, probably for @PatReynolds: Can we illustrate these computer vision projects by using a handful (1-6) of images from the Census records and parish registers? It would be a lot more compelling if we could, but I don't know about the rights situation.
Thanks Ben. @PatReynolds do we have any other potential projects after our discussions with the CSi folks?
CSi was about machine markup of images. The other AI project is machine learning of effective ways of searching FreeCEN data.
FreeCEN2 Issue 411 is the other AI project
Pat's story involves using a data distribution to suggest ways to narrow down their query as they enter it.
A separate story is to suggest linkages between the same individuals between different census years. We might be able to get "gold standard" data from Oliver Duke-Williams
Another idea involves OCRing will images, running NLP for Named Entity Recognition to index them.
Similar to will images, the Catholic Record Society texts could be auto-indexed through NER.
As per Ben's example relating to using AI (computer vision) for helping with detecting census forms. Can we come up with a list of such example projects for new volunteers? AI/machine learning and deep learning are areas of tech many people are looking to gain experience in..