Open ashwath1295 opened 2 years ago
One way is to do custom training. Get some 50-100 images, label them using a tool like labelImg, (https://github.com/tzutalin/labelImg.git), then train your model and get the weights. You can then evaluate any image or video, and the output will only have one class.
I want to evaluate the Yolov4 model just for persons. How can I do that?