Towards a Better Understanding and Regularization of GAN Training Dynamics

Abstract

Generative adversarial networks (GANs) are notoriously difficult to train and the reasons underlying their (non-)convergence behaviors are still not completely understood. By first considering a simple yet representative GAN example, we mathematically analyze its local convergence behavior in a non-asymptotic way. Furthermore, the analysis is extended to general GANs under certain assumptions. We find that in order to ensure a good convergence rate, two factors of the Jacobian in the GAN training dynamics should be simultaneously avoided, which are (i) the Phase Factor, i.e., the Jacobian has complex eigenvalues with a large imaginary-to-real ratio, and (ii) the Conditioning Factor, i.e., the Jacobian is ill-conditioned. Previous methods of regularizing the Jacobian can only alleviate one of these two factors, while making the other more severe. Thus we propose a new JAcobian REgularization (JARE) for GANs, which simultaneously addresses both factors by construction. Finally, we conduct experiments that confirm our theoretical analysis and demonstrate the advantages of JARE over previous methods in stabilizing GANs.

arXiv commentUAI 2019

References in corpus (12)

• Spectral Normalization for Generative Adversarial Networks
• Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
• Improved Training of Wasserstein GANs
• Improved Techniques for Training GANs
• Which Training Methods for GANs do actually Converge?
• f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization
• Generative Adversarial Imitation Learning
• Stabilizing Training of Generative Adversarial Networks through Regularization
• A Variational Inequality Perspective on Generative Adversarial Networks
• The Mechanics of n-Player Differentiable Games
• Gradient descent GAN optimization is locally stable
• The Limit Points of (Optimistic) Gradient Descent in Min-Max Optimization

Cited by in corpus (6)

• Convergence of Learning Dynamics in Stackelberg Games
• Statistical guarantees for generative models without domination
• Learning Flat Latent Manifolds with VAEs
• A Novel Framework for Selection of GANs for an Application
• Online Kernel based Generative Adversarial Networks
• Beyond Folklore: A Scaling Calculus for the Design and Initialization of ReLU Networks