Lightweight 3D Gaussian Splatting Compression via Video Codec

TL;DR

LGSCV tackles the heavy computational burden of video-based 3D Gaussian Splatting compression by replacing PLAS with a lightweight two-stage Morton scan to create blockwise 2D maps, applying PCA to compress SH AC components, and using a fast MiniPLAS within small blocks. This combination yields significant rate–distortion gains (over 20% on common datasets) and markedly reduces map-generation and encoding times, enabling GS compression on lightweight devices. The approach also provides practical guidance for codec unit sizing and demonstrates robust gains across medium-to-low bitrates, validated against the MPEG GS compression benchmark. Overall, LGSCV offers a practical, faster, and RD-efficient pathway for compressing 3D GS data with standard video codecs.

Abstract

Current video-based GS compression methods rely on using Parallel Linear Assignment Sorting (PLAS) to convert 3D GS into smooth 2D maps, which are computationally expensive and time-consuming, limiting the application of GS on lightweight devices. In this paper, we propose a Lightweight 3D Gaussian Splatting (GS) Compression method based on Video codec (LGSCV). First, a two-stage Morton scan is proposed to generate blockwise 2D maps that are friendly for canonical video codecs in which the coding units (CU) are square blocks. A 3D Morton scan is used to permute GS primitives, followed by a 2D Morton scan to map the ordered GS primitives to 2D maps in a blockwise style. However, although the blockwise 2D maps report close performance to the PLAS map in high-bitrate regions, they show a quality collapse at medium-to-low bitrates. Therefore, a principal component analysis (PCA) is used to reduce the dimensionality of spherical harmonics (SH), and a MiniPLAS, which is flexible and fast, is designed to permute the primitives within certain block sizes. Incorporating SH PCA and MiniPLAS leads to a significant gain in rate-distortion (RD) performance, especially at medium and low bitrates. MiniPLAS can also guide the setting of the codec CU size configuration and significantly reduce encoding time. Experimental results on the MPEG dataset demonstrate that the proposed LGSCV achieves over 20% RD gain compared with state-of-the-art methods, while reducing 2D map generation time to approximately 1 second and cutting encoding time by 50%. The code is available at https://github.com/Qi-Yangsjtu/LGSCV .

Figures (4)

• Figure 1: Framework of the proposed LGSCV.
• Figure 2: Performance comparison on MPEG dataset.
• Figure 3: Ablation study of LGSCV.
• Figure 4: Ablation Study on CU