yeephycho/tensorflow-face-detection

2 classes defined instead of just one

eypros opened this issue · 1 comments

I would like to ask for the 2 classes that you have defined. Is there a specific reason for that?

I have tried to train a one-class with persons and it seemed to me that sometimes multiple classes gives a precision boost for some reason. Is this the case here? Because as far as I can see you don't actually uses the background class at all. Are there any background class member that the model could detect for example?

I don't think so, if you found that a single class doesn't works out, it may because of other issue.