In DCGAN, dropout is not used in either generator or discriminator.
When using batch norm, are the benefits of dropout generally so marginal that is is not used?
If it is used, in what circumstances? Both discriminator and generator?
Asked
Active
Viewed 578 times
6
-
using both hurts performance -- dropout increases variance at train time, breaking the batch norm assumption that activation statistics should be roughly the same from train to test – shimao Apr 09 '19 at 21:12