Question/Issue: Is it possible to use multithreading for FOMO and OCR, with OCR being independent of FOMO?
Yes you can, just download the C++ library and then you can do the inferencing in multithreads.
We have also example on how to use the SDK on our website.
However I do not see too many use cases, unless if you have multiple sensors and each one has its own inferencing thread so this can work.
Also you can use OCR with FOMO but why you would need this?