Pose-free 3D Gaussian splatting via shape-ray estimation

Summary

[Submitted on 29 May 2025 (v1), last revised 21 Oct 2025 (this version, v3)] Title:Pose-free 3D Gaussian splatting via shape-ray estimation View a PDF of the paper titled Pose-free 3D Gaussian splatting via shape-ray estimation, by Youngju Na and 5 other authors View PDF HTML (experimental) Abstract:While generalizable 3D Gaussian splatting enables efficient, high-quality rendering of unseen scenes, it heavily depends on precise camera poses for accurate geometry. In real-world scenarios, obtaining accurate poses is challenging, leading to noisy pose estimates and geometric misalignments. To address this, we introduce SHARE, a pose-free, feed-forward Gaussian splatting framework that overcomes these ambiguities by joint shape and camera rays estimation. Instead of relying on explicit 3D transformations, SHARE builds a pose-aware canonical volume representation that seamlessly integrates multi-view information, reducing misalignment caused by inaccurate pose estimates. Additionally, anchor-aligned Gaussian prediction enhances scene reconstruction by refining local geometry around coarse anchors, allowing for more precise Gaussian placement. Extensive experiments on diverse real-world datasets show that our method achieves robust performance in pose-free generalizable Gaussian splatting. Code is avilable at this https URL Submission history From: Youngju Na [view email] [v1] Thu, 29 May 2025 01:34:40 UTC (1,331 KB) [v2] Fri, 26 Sep 2025 06:08:53 UTC (1,331 KB) [v3] Tue, 21 Oct 2025 11:48:43 UTC (1,331 KB)