GAN-based Image Compression with Improved RDO Process

TL;DR

This paper introduces a GAN-based image compression method that enhances perceptual quality and entropy modeling using improved RDO, DISTS, MS-SSIM, and GLLMM, validated by MOS experiments showing superior performance.

Contribution

It proposes a novel GAN-based compression approach with an improved RDO process, incorporating advanced perceptual metrics and entropy modeling for better image quality.

Findings

Outperforms existing GAN-based methods

Achieves higher perceptual quality in MOS tests

Surpasses state-of-the-art VVC codec

Abstract

GAN-based image compression schemes have shown remarkable progress lately due to their high perceptual quality at low bit rates. However, there are two main issues, including 1) the reconstructed image perceptual degeneration in color, texture, and structure as well as 2) the inaccurate entropy model. In this paper, we present a novel GAN-based image compression approach with improved rate-distortion optimization (RDO) process. To achieve this, we utilize the DISTS and MS-SSIM metrics to measure perceptual degeneration in color, texture, and structure. Besides, we absorb the discretized gaussian-laplacian-logistic mixture model (GLLMM) for entropy modeling to improve the accuracy in estimating the probability distributions of the latent representation. During the evaluation process, instead of evaluating the perceptual quality of the reconstructed image via IQA metrics, we directly conduct the Mean Opinion Score (MOS) experiment among different codecs, which fully reflects the actual perceptual results of humans. Experimental results demonstrate that the proposed method outperforms the existing GAN-based methods and the state-of-the-art hybrid codec (i.e., VVC).