Hi, Thanks for your beautiful work, i have a question about paper:
The reason that you didn't take global discriminator is that the recptive field is already satisify the training input size , so you mean the size of the image in the training set is smaller than the input to the discriminator? otherwise , if they are same size , the output is only one value but not a matirx. （because the recptive field is the size of training set images)
diaodeyi
commented
3 years ago
Hi, Thanks for your beautiful work, i have a question about paper: The reason that you didn't take global discriminator is that the recptive field is already satisify the training input size , so you mean the size of the image in the training set is smaller than the input to the discriminator? otherwise , if they are same size , the output is only one value but not a matirx. （because the recptive field is the size of training set images)