I use the origin yolov3 weights to predict my own dataset, and the prediction result is not bad. However, the speed of the prediction can not meet the requirement.
My task is about detecting cars and counting them from the CCTV on the road, so i do not care about the other object. Maybe it can be a binary classification, detecting an object is a car or not.
So is the any solution to speed up the prediction?
I use the origin yolov3 weights to predict my own dataset, and the prediction result is not bad. However, the speed of the prediction can not meet the requirement.
My task is about detecting cars and counting them from the CCTV on the road, so i do not care about the other object. Maybe it can be a binary classification, detecting an object is a car or not.
So is the any solution to speed up the prediction?
Thanks All :)