GaussianPOP: Principled Simplification Framework for Compact 3D Gaussian Splatting via Error Quantification

TL;DR

GaussianPOP introduces a principled, error-based simplification framework for 3D Gaussian Splatting that improves compactness without sacrificing rendering quality, outperforming existing methods.

Contribution

A novel error criterion derived from the 3DGS rendering equation enables accurate Gaussian simplification with an efficient algorithm supporting both training and post-training pruning.

Findings

Outperforms state-of-the-art pruning methods in compactness and quality.

Supports efficient error calculation in a single forward pass.

Provides flexible pruning strategies for different application needs.

Abstract

Existing 3D Gaussian Splatting simplification methods commonly use importance scores, such as blending weights or sensitivity, to identify redundant Gaussians. However, these scores are not driven by visual error metrics, often leading to suboptimal trade-offs between compactness and rendering fidelity. We present GaussianPOP, a principled simplification framework based on analytical Gaussian error quantification. Our key contribution is a novel error criterion, derived directly from the 3DGS rendering equation, that precisely measures each Gaussian's contribution to the rendered image. By introducing a highly efficient algorithm, our framework enables practical error calculation in a single forward pass. The framework is both accurate and flexible, supporting on-training pruning as well as post-training simplification via iterative error re-quantification for improved stability. Experimental results show that our method consistently outperforms existing state-of-the-art pruning methods across both application scenarios, achieving a superior trade-off between model compactness and high rendering quality.