A Benchmark for Gaussian Splatting Compression and Quality Assessment Study

TL;DR

This work addresses the gap in traditional compression for 3D Gaussian Splatting (GS) by proposing Graph-based GS Compression (GGSC), a simple yet effective anchor that splits GS with KDTree, builds local graphs, and applies Graph Fourier Transform-based residual clipping before quantization and coding. It is complemented by GSQA, a large-scale subjective dataset with static and dynamic GS content to study how both high-frequency clipping and quantization distort GS attributes and affect perceived quality. The study evaluates existing objective metrics on GSQA, revealing that while some metrics correlate well with MOS, there is a clear need for GS-specific quality measures and smarter bitrate- distortion trade-offs across attributes. Overall, the paper provides open-source GGSC and GSQA resources and outlines future directions for optimal GS parameterization and compatibility with other GS-derived representations.

Abstract

To fill the gap of traditional GS compression method, in this paper, we first propose a simple and effective GS data compression anchor called Graph-based GS Compression (GGSC). GGSC is inspired by graph signal processing theory and uses two branches to compress the primitive center and attributes. We split the whole GS sample via KDTree and clip the high-frequency components after the graph Fourier transform. Followed by quantization, G-PCC and adaptive arithmetic coding are used to compress the primitive center and attribute residual matrix to generate the bitrate file. GGSS is the first work to explore traditional GS compression, with advantages that can reveal the GS distortion characteristics corresponding to typical compression operation, such as high-frequency clipping and quantization. Second, based on GGSC, we create a GS Quality Assessment dataset (GSQA) with 120 samples. A subjective experiment is conducted in a laboratory environment to collect subjective scores after rendering GS into Processed Video Sequences (PVS). We analyze the characteristics of different GS distortions based on Mean Opinion Scores (MOS), demonstrating the sensitivity of different attributes distortion to visual quality. The GGSC code and the dataset, including GS samples, MOS, and PVS, are made publicly available at https://github.com/Qi-Yangsjtu/GGSC.

Figures (10)

• Figure 1: End-to-end GS compression system.
• Figure 2: Flowchart of GGSC.
• Figure 3: Snapshot of GSD source synthetic objects.
• Figure 4: Snapshot of GSD source unbounded photo-realistic scenes and dynamic content.
• Figure 5: SI vs. TI.