Training Generative Adversarial Networks via Primal-Dual subgradient Methods: a Lagrangian Perspective on GaN.

Xu Chen, Jiang Wang, Hao Ge

Browse the full ICLR paper archive.