YOLO Loss Function

Hi all,

I just want to confirm that Professor Ng did not go through the actual formulation of the loss function for the YOLO algorithm right?

If so, can we just assume that the loss function is the same/similar to that of object localization with sum of squared-error for the various label outputs, then applied for every grid cell and every anchor box? See below for his formulation of the object localization algorithm loss:

Thank you!


That’s from the original 2016 paper