ConeGS: Error-Guided Densification Using Pixel Cones for Improved Reconstruction With Fewer Primitives

TL;DR

ConeGS introduces an error-guided, pixel cone-based densification method for 3D Gaussian Splatting that improves scene reconstruction quality and efficiency, especially under primitive constraints, by intelligently placing primitives along viewing cones.

Contribution

The paper presents ConeGS, a novel densification framework that uses pixel cones and scene-aware error estimation to optimize primitive placement independently of existing geometry.

Findings

Significantly improves reconstruction quality across various primitive budgets.

Enhances rendering performance, especially under tight primitive constraints.

Effectively reduces the number of primitives needed for high-quality scene representation.

Abstract

3D Gaussian Splatting (3DGS) achieves state-of-the-art image quality and real-time performance in novel view synthesis but often suffers from a suboptimal spatial distribution of primitives. This issue stems from cloning-based densification, which propagates Gaussians along existing geometry, limiting exploration and requiring many primitives to adequately cover the scene. We present ConeGS, an image-space-informed densification framework that is independent of existing scene geometry state. ConeGS first creates a fast Instant Neural Graphics Primitives (iNGP) reconstruction as a geometric proxy to estimate per-pixel depth. During the subsequent 3DGS optimization, it identifies high-error pixels and inserts new Gaussians along the corresponding viewing cones at the predicted depth values, initializing their size according to the cone diameter. A pre-activation opacity penalty rapidly removes redundant Gaussians, while a primitive budgeting strategy controls the total number of primitives, either by a fixed budget or by adapting to scene complexity, ensuring high reconstruction quality. Experiments show that ConeGS consistently enhances reconstruction quality and rendering performance across Gaussian budgets, with especially strong gains under tight primitive constraints where efficient placement is crucial.