dataset

how many images did you use to train the model.
i used approx 500 images for 1 class but still the output is not correct.
Like it is not showing bounding boxes for all the images and in some cases wrong,
what could be the possible error?