Adversarial Training Improves Joint Energy-Based Generative Modelling

TL;DR

This paper introduces a hybrid energy-based generative model that leverages adversarial training, combining interpretable gradients and Langevin Dynamics to enhance stability, robustness, and generative quality.

Contribution

It presents a novel framework integrating adversarial training with hybrid energy-based models, improving training stability and generative performance.

Findings

Enhanced training stability and robustness.

Improved generative modeling quality.

Effective combination of gradients and Langevin Dynamics.

Abstract

We propose the novel framework for generative modelling using hybrid energy-based models. In our method we combine the interpretable input gradients of the robust classifier and Langevin Dynamics for sampling. Using the adversarial training we improve not only the training stability, but robustness and generative modelling of the joint energy-based models.