Adversarial confidence and smoothness regularizations for scalable unsupervised discriminative learning

Re-assign community