Steepest Descent Density Control for Compact 3D Gaussian Splatting

TL;DR

This paper introduces SteepGS, a novel density control method for 3D Gaussian Splatting that reduces point cloud size by about 50% while preserving rendering quality, enabling more efficient and scalable scene representations.

Contribution

It provides a theoretical framework and optimization approach for density control in 3D Gaussian Splatting, leading to a compact point cloud with maintained quality.

Findings

Achieves ~50% reduction in Gaussian points

Maintains high-quality rendering with fewer points

Enhances efficiency and scalability of 3D scene representations

Abstract

3D Gaussian Splatting (3DGS) has emerged as a powerful technique for real-time, high-resolution novel view synthesis. By representing scenes as a mixture of Gaussian primitives, 3DGS leverages GPU rasterization pipelines for efficient rendering and reconstruction. To optimize scene coverage and capture fine details, 3DGS employs a densification algorithm to generate additional points. However, this process often leads to redundant point clouds, resulting in excessive memory usage, slower performance, and substantial storage demands - posing significant challenges for deployment on resource-constrained devices. To address this limitation, we propose a theoretical framework that demystifies and improves density control in 3DGS. Our analysis reveals that splitting is crucial for escaping saddle points. Through an optimization-theoretic approach, we establish the necessary conditions for densification, determine the minimal number of offspring Gaussians, identify the optimal parameter update direction, and provide an analytical solution for normalizing off-spring opacity. Building on these insights, we introduce SteepGS, incorporating steepest density control, a principled strategy that minimizes loss while maintaining a compact point cloud. SteepGS achieves a ~50% reduction in Gaussian points without compromising rendering quality, significantly enhancing both efficiency and scalability.