https://learnopencv.com/deep-learning-with-opencvs-dnn-module-a-definitive-guide/ After training we can convert some specific networks into OPencv DNN compatable please follow the link to convert tensorflow model into Dnn readable . What is the need for convertion ? Usually tensorflow models require protobuf installation which is little bit complex framework and runs slowly on cpu where as with opencv Dnn module we can achive decent amount of FPS even with out GPU(if it is SSD we achieve 22 fps and with YOLO V3 and V4 it should be around 7-9 fps). https://github.com/opencv/opencv/wiki/TensorFlow-Object-Detection-API
Results woulbe as below on Kitty data set sequence 7
