FilterGS: Traversal-Free Parallel Filtering and Adaptive Shrinking for Large-Scale LoD 3D Gaussian Splatting

TL;DR

FilterGS introduces a traversal-free, parallel filtering and adaptive shrinking method for large-scale 3D Gaussian Splatting, significantly improving rendering speed while preserving visual quality.

Contribution

It presents a novel parallel filtering mechanism and a scene-adaptive Gaussian shrinking strategy that eliminate the need for tree traversal and reduce redundancy in large-scale 3D Gaussian rendering.

Findings

Achieves state-of-the-art rendering speeds on large datasets.

Maintains competitive visual quality with reduced processing overhead.

Effectively reduces redundant Gaussian-tile pairs using the GTC metric.

Abstract

3D Gaussian Splatting has revolutionized neural rendering with real-time performance. However, scaling this approach to large scenes using Level-of-Detail methods faces critical challenges: inefficient serial traversal consuming over 60\% of rendering time, and redundant Gaussian-tile pairs that incur unnecessary processing overhead. To address these limitations, we introduce FilterGS, featuring a parallel filtering mechanism with two complementary filters that select Gaussian elements efficiently without tree traversal. Additionally, we propose a novel GTC metric that quantifies the redundancy of Gaussian-tile key-value pairs. Based on this metric, we introduce a scene-adaptive Gaussian shrinking strategy that effectively reduces redundant pairs. Extensive experiments demonstrate that FilterGS achieves state-of-the-art rendering speeds while maintaining competitive visual quality across multiple large-scale datasets. Project page: https://github.com/xenon-w/FilterGS