It doesn’t matter. Even the training process makes it pretty much impossible to tell these things apart.
And if we do find a way to distinguish, we’ll immediately incorporate that into the model design in a GAN like manner, and we’ll soon be unable to distinguish again.
The findings were for GAN models, not GAN like components though.
It doesn’t matter. Even the training process makes it pretty much impossible to tell these things apart.
And if we do find a way to distinguish, we’ll immediately incorporate that into the model design in a GAN like manner, and we’ll soon be unable to distinguish again.