GFix: Perceptually Enhanced Gaussian Splatting Video Compression

TL;DR

GFix introduces a perceptually enhanced, content-adaptive neural framework for 3D Gaussian Splatting video compression, significantly improving visual quality and compression efficiency by leveraging diffusion models and a novel modulated LoRA scheme.

Contribution

The paper proposes GFix, a novel neural enhancement framework that improves 3DGS video compression through perceptual enhancement and efficient diffusion model adaptation.

Findings

Achieves up to 72.1% BD-rate savings in LPIPS.

Outperforms GSVC in perceptual quality metrics.

Demonstrates significant visual quality improvements.

Abstract

3D Gaussian Splatting (3DGS) enhances 3D scene reconstruction through explicit representation and fast rendering, demonstrating potential benefits for various low-level vision tasks, including video compression. However, existing 3DGS-based video codecs generally exhibit more noticeable visual artifacts and relatively low compression ratios. In this paper, we specifically target the perceptual enhancement of 3DGS-based video compression, based on the assumption that artifacts from 3DGS rendering and quantization resemble noisy latents sampled during diffusion training. Building on this premise, we propose a content-adaptive framework, GFix, comprising a streamlined, single-step diffusion model that serves as an off-the-shelf neural enhancer. Moreover, to increase compression efficiency, We propose a modulated LoRA scheme that freezes the low-rank decompositions and modulates the intermediate hidden states, thereby achieving efficient adaptation of the diffusion backbone with highly compressible updates. Experimental results show that GFix delivers strong perceptual quality enhancement, outperforming GSVC with up to 72.1% BD-rate savings in LPIPS and 21.4% in FID.