Light3R-SfM: Towards Feed-forward Structure-from-Motion

TL;DR

Light3R-SfM introduces a fast, learnable structure-from-motion framework that replaces traditional optimization with attention mechanisms, enabling scalable and efficient 3D reconstruction from large image collections.

Contribution

It proposes a novel feed-forward SfM method with a global alignment module and scene graph construction, reducing computational costs while maintaining accuracy.

Findings

Achieves competitive accuracy with reduced runtime

Uses a learnable attention-based global alignment module

Constructs a sparse scene graph for efficiency

Abstract

We present Light3R-SfM, a feed-forward, end-to-end learnable framework for efficient large-scale Structure-from-Motion (SfM) from unconstrained image collections. Unlike existing SfM solutions that rely on costly matching and global optimization to achieve accurate 3D reconstructions, Light3R-SfM addresses this limitation through a novel latent global alignment module. This module replaces traditional global optimization with a learnable attention mechanism, effectively capturing multi-view constraints across images for robust and precise camera pose estimation. Light3R-SfM constructs a sparse scene graph via retrieval-score-guided shortest path tree to dramatically reduce memory usage and computational overhead compared to the naive approach. Extensive experiments demonstrate that Light3R-SfM achieves competitive accuracy while significantly reducing runtime, making it ideal for 3D reconstruction tasks in real-world applications with a runtime constraint. This work pioneers a data-driven, feed-forward SfM approach, paving the way toward scalable, accurate, and efficient 3D reconstruction in the wild.