Kernel-Guided Training of Implicit Generative Models with Stability Guarantees

TL;DR

This paper introduces a kernel-guided regularization method for GANs that enhances training stability and interpretability by controlling distribution discrepancies, backed by theoretical guarantees and diverse experiments.

Contribution

It proposes a novel kernel-based regularization approach to stabilize GAN training with theoretical stability guarantees.

Findings

Improved training stability of GANs.

Enhanced interpretability of generative models.

Theoretical proof of stability guarantees.

Abstract

Modern implicit generative models such as generative adversarial networks (GANs) are generally known to suffer from issues such as instability, uninterpretability, and difficulty in assessing their performance. If we see these implicit models as dynamical systems, some of these issues are caused by being unable to control their behavior in a meaningful way during the course of training. In this work, we propose a theoretically grounded method to guide the training trajectories of GANs by augmenting the GAN loss function with a kernel-based regularization term that controls local and global discrepancies between the model and true distributions. This control signal allows us to inject prior knowledge into the model. We provide theoretical guarantees on the stability of the resulting dynamical system and demonstrate different aspects of it via a wide range of experiments.