Training Implicit Generative Models via an Invariant Statistical Loss

Open in new window