Dimension of discriminator output in pix2pixBEGAN.py

[alliecc]

Hello,

Thanks for sharing codes. I was just wondering why you don't make BEGAN discriminator condition on input?

[taey16]

@alliecc ,
I thought that
BEGAN discriminator is an auto-encoder so that input images should be targets with mean-squared error (MSE) loss function. Conditioning input (as did in pix2pix paper) isn't make sense because BEGAN discriminator is not a classifier but an auto-encoder.