RJCode

my question is when we get the inference output, will we get the accuracy of that output or not?

In terms of MOTA (multiple object tracking accuracy) during demo inference without training.