Neural Video Compression using GANs for Detail Synthesis and Propagation

TL;DR

This paper introduces a neural video compression method based on GANs that significantly improves visual quality by synthesizing and propagating details, outperforming previous methods in user studies.

Contribution

The paper presents the first GAN-based neural video compression approach that effectively synthesizes and propagates details, setting a new state-of-the-art in visual quality.

Findings

GAN loss is crucial for high visual quality

User studies are necessary for fair comparison

The method outperforms previous neural and non-neural approaches

Abstract

We present the first neural video compression method based on generative adversarial networks (GANs). Our approach significantly outperforms previous neural and non-neural video compression methods in a user study, setting a new state-of-the-art in visual quality for neural methods. We show that the GAN loss is crucial to obtain this high visual quality. Two components make the GAN loss effective: we i) synthesize detail by conditioning the generator on a latent extracted from the warped previous reconstruction to then ii) propagate this detail with high-quality flow. We find that user studies are required to compare methods, i.e., none of our quantitative metrics were able to predict all studies. We present the network design choices in detail, and ablate them with user studies.