Saved in:
Bibliographic Details
Main Authors: Hofer, Manuel, Steinberger, Markus, Köhler, Thomas
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.03327
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917244519317504
author Hofer, Manuel
Steinberger, Markus
Köhler, Thomas
author_facet Hofer, Manuel
Steinberger, Markus
Köhler, Thomas
contents Novel view synthesis has evolved rapidly, advancing from Neural Radiance Fields to 3D Gaussian Splatting (3DGS), which offers real-time rendering and rapid training without compromising visual fidelity. However, 3DGS relies heavily on accurate camera poses and high-quality point cloud initialization, which are difficult to obtain in sparse-view scenarios. While traditional Structure from Motion (SfM) pipelines often fail in these settings, existing learning-based point estimation alternatives typically require reliable reference views and remain sensitive to pose or depth errors. In this work, we propose a robust method utilizing π^3, a reference-free point cloud estimation network. We integrate dense initialization from π^3 with a regularization scheme designed to mitigate geometric inaccuracies. Specifically, we employ uncertainty-guided depth supervision, normal consistency loss, and depth warping. Experimental results demonstrate that our approach achieves state-of-the-art performance on the Tanks and Temples, LLFF, DTU, and MipNeRF360 datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2602_03327
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Pi-GS: Sparse-View Gaussian Splatting with Dense π^3 Initialization
Hofer, Manuel
Steinberger, Markus
Köhler, Thomas
Graphics
Computer Vision and Pattern Recognition
Novel view synthesis has evolved rapidly, advancing from Neural Radiance Fields to 3D Gaussian Splatting (3DGS), which offers real-time rendering and rapid training without compromising visual fidelity. However, 3DGS relies heavily on accurate camera poses and high-quality point cloud initialization, which are difficult to obtain in sparse-view scenarios. While traditional Structure from Motion (SfM) pipelines often fail in these settings, existing learning-based point estimation alternatives typically require reliable reference views and remain sensitive to pose or depth errors. In this work, we propose a robust method utilizing π^3, a reference-free point cloud estimation network. We integrate dense initialization from π^3 with a regularization scheme designed to mitigate geometric inaccuracies. Specifically, we employ uncertainty-guided depth supervision, normal consistency loss, and depth warping. Experimental results demonstrate that our approach achieves state-of-the-art performance on the Tanks and Temples, LLFF, DTU, and MipNeRF360 datasets.
title Pi-GS: Sparse-View Gaussian Splatting with Dense π^3 Initialization
topic Graphics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2602.03327