Implement Object Detection in Deep ROS

[lucasreljic]

Transfer and revamp camera object detection for N number of cameras running TensorRT for inference only in the Deep ROS repository. Then it must be transferred or linked to the monorepo.

This includes configurations for traffic lights and traffic sign models.

Inferencing Node should be written in C++