ML/AI recognition of cyrillic documents
Hi everyone. I'm an independent ML/AI developer and I would like to develop recognition algorithm for cyrillic documents (or participate in the project as contributor if there is something ongoing) as my part of Involvement to the great work of FamilySearch. I believe although it would be difficult but definitely doable. I would appreciate a lot if I can discuss the project with someone from FamilySearch team.
Answers
-
You would have to go through regular channels at the Family History Center in Salt Lake. We would have no idea here as to how to go about doing that. Good luck with your project!
0 -
I don't know if this is related to the project that Evgenii was refering to, but recently I was at Family Home Evening event for a Young Single Adult ward where we attempted to index certain documents, but found it very hard to interpret the names. A friend of mine who's experienced with using AI, took a screen shot of some of the names on the documents and put it into an AI chat box. He was able to get a few results that gave us a better idea of the name. Would it be possible to implement a kind of chat box within Family Search to do that to provide suggestions for names for us to consider as we index?
1
