Closed FormatFish closed 7 years ago
You can see that both generator
and discriminator
model include y
term if dataset is mnist
. I think y_dim
means number of dimension of label of each data (I mean, if dataset is mnist
, it means 10. e.g. 0, 1, 2, ... 9.) and this model architecture is closer to cGAN rather than LAPGAN.
Yes it's a conditional variable which is described in appendix of the paper which makes DCGAN as a conditional generative model.
so
Hi , I just read the paper and your code , and I find "y_dim" in your model.py function of generator , Is that the label for vector z? and just like "LAPGAN" , the y_dim is just like conditioning variable which aim is to control the generator model?