My YoloV7 Model Predicts way too many objects Than expected

Viewed 25

I have successfully trained my Yolov7 model, but after running the detect.py script, the displayed outcome is totally awkward. The funny thing is I've trained on a different dataset and got a pretty decent result, but I've tried again with about 5 more datasets and the results I get are troubling. the model displays like a million bounding boxes not related to what I trained it for. I doubt if it's the dataset.

screenshot of the output.

enter image description here

0 Answers
Related