LongSplat: Robust Unposed 3D Gaussian Splatting for Casual Long Videos

TL;DR

LongSplat introduces a robust framework for novel view synthesis from long, casually captured videos, effectively handling irregular camera motion and large scenes through joint optimization, learned priors, and efficient data structures.

Contribution

It presents a novel unposed 3D Gaussian Splatting method with incremental joint optimization, a learned pose estimation module, and an octree-based anchor formation, addressing key challenges in long video view synthesis.

Findings

Achieves state-of-the-art rendering quality on challenging benchmarks.

Improves pose accuracy and computational efficiency over previous methods.

Effectively handles irregular camera motion and large scenes.

Abstract

LongSplat addresses critical challenges in novel view synthesis (NVS) from casually captured long videos characterized by irregular camera motion, unknown camera poses, and expansive scenes. Current methods often suffer from pose drift, inaccurate geometry initialization, and severe memory limitations. To address these issues, we introduce LongSplat, a robust unposed 3D Gaussian Splatting framework featuring: (1) Incremental Joint Optimization that concurrently optimizes camera poses and 3D Gaussians to avoid local minima and ensure global consistency; (2) a robust Pose Estimation Module leveraging learned 3D priors; and (3) an efficient Octree Anchor Formation mechanism that converts dense point clouds into anchors based on spatial density. Extensive experiments on challenging benchmarks demonstrate that LongSplat achieves state-of-the-art results, substantially improving rendering quality, pose accuracy, and computational efficiency compared to prior approaches. Project page: https://linjohnss.github.io/longsplat/