Open paranoidai opened 6 years ago
I've tried to train the model on my own dataset but when testing, the model will detect some obj for every picture, even there is no obj in the picture, should I add some negative data which contain no object windows? if so ,how should I use them?
I have the same issue with. I am wondering if the yolo_loss function allow an empty box list?
I've tried to train the model on my own dataset but when testing, the model will detect some obj for every picture, even there is no obj in the picture, should I add some negative data which contain no object windows? if so ,how should I use them?