Generate High Resolution Images With Generative Variational Autoencoder

TL;DR

This paper introduces a novel neural network architecture that enhances high-resolution image generation by integrating a discriminator into a variational autoencoder, achieving superior quality and sharper images across multiple datasets.

Contribution

The work proposes replacing the VAE decoder with a discriminator and combining encoder and generator inputs, resulting in improved high-resolution image generation with better evaluation metrics.

Findings

Outperforms previous state-of-the-art on multiple datasets

Generates sharper and higher quality images

Achieves better metrics like MMD, SSIM, and ELBO

Abstract

In this work, we present a novel neural network to generate high resolution images. We replace the decoder of VAE with a discriminator while using the encoder as it is. The encoder is fed data from a normal distribution while the generator is fed from a gaussian distribution. The combination from both is given to a discriminator which tells whether the generated image is correct or not. We evaluate our network on 3 different datasets: MNIST, LSUN and CelebA dataset. Our network beats the previous state of the art using MMD, SSIM, log likelihood, reconstruction error, ELBO and KL divergence as the evaluation metrics while generating much sharper images. This work is potentially very exciting as we are able to combine the advantages of generative models and inference models in a principled bayesian manner.