A Bayesian Non-parametric Approach to Generative Models: Integrating Variational Autoencoder and Generative Adversarial Networks using Wasserstein and Maximum Mean Discrepancy

TL;DR

This paper introduces a Bayesian non-parametric framework that combines Wasserstein and MMD measures with GANs, VAEs, and a code-GAN to improve sample quality, diversity, and training stability in generative modeling.

Contribution

It presents a novel BNPL-based triple model integrating GAN, VAE, and CGAN with WMMD loss, enhancing robustness and performance in generative tasks.

Findings

Improved training stability and robustness demonstrated.

High-quality, diverse samples achieved across datasets.

Theoretical and empirical validation supports effectiveness.

Abstract

We propose a novel generative model within the Bayesian non-parametric learning (BNPL) framework to address some notable failure modes in generative adversarial networks (GANs) and variational autoencoders (VAEs)--these being overfitting in the GAN case and noisy samples in the VAE case. We will demonstrate that the BNPL framework enhances training stability and provides robustness and accuracy guarantees when incorporating the Wasserstein distance and maximum mean discrepancy measure (WMMD) into our model's loss function. Moreover, we introduce a so-called ``triple model'' that combines the GAN, the VAE, and further incorporates a code-GAN (CGAN) to explore the latent space of the VAE. This triple model design generates high-quality, diverse samples, while the BNPL framework, leveraging the WMMD loss function, enhances training stability. Together, these components enable our model to achieve superior performance across various generative tasks. These claims are supported by both theoretical analyses and empirical validation on a wide variety of datasets.