Exploiting the Hidden Tasks of GANs: Making Implicit Subproblems Explicit

Abstract

We present an alternative perspective on the training of generative adversarial networks (GANs), showing that the training step for a GAN generator decomposes into two implicit subproblems. In the first, the discriminator provides new target data to the generator in the form of "inverse examples" produced by approximately inverting classifier labels. In the second, these examples are used as targets to update the generator via least-squares regression, regardless of the main loss specified to train the network. We experimentally validate our main theoretical result and demonstrate significant improvements over standard GAN training made possible by making these subproblems explicit.

arXiv comment12 pages, 3 figures

References in corpus (9)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• NIPS 2016 Tutorial: Generative Adversarial Networks
• A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
• Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step
• Implicit Maximum Likelihood Estimation
• Global Convergence to the Equilibrium of GANs using Variational Inequalities
• On Solving Minimax Optimization Locally: A Follow-the-Ridge Approach
• On the Implicit Assumptions of GANs