LInux and C++ based Multi camera tracking using Computer vision and Deep Learning - Rework

There is a repo application for decoding multiple HD video streams on an Intel processor on a Linux based system written in C++ and Opencv libs using intel media libraries which makes it optimal in performance. This application needs to worked on so that information (AI Inference results can be exposed without breaking the throughput we are getting ) can be passed to a higher level routine so that higher level processing can be carried out