I am currently facing the issue, that the model is drawing multiple boxes around the same image, even though the object is not in the picture, something like this, where e.g.:
Blue = Cat
Red = Dog
But both are on the cat. Do you have an idea how to prevent this from happening? Is it maybe possible to give a minimum threshold of confidence for boxes to be considered and drawn?
This shows that the classification of the detector is not good enough to distinguish these classes. In your example, the classification of cats and dogs is a challenging task.
Hello,
I am currently facing the issue, that the model is drawing multiple boxes around the same image, even though the object is not in the picture, something like this, where e.g.:
Blue = Cat Red = Dog
But both are on the cat. Do you have an idea how to prevent this from happening? Is it maybe possible to give a minimum threshold of confidence for boxes to be considered and drawn?
Thanks and best regards