Global-Aware Edge Prioritization for Pose Graph Initialization

TL;DR

This paper introduces a global-aware edge prioritization method for pose graph initialization in SfM, using a GNN to rank candidate edges based on global consistency, leading to more reliable and compact graphs.

Contribution

It proposes a novel GNN-based approach for globally consistent edge ranking, improving pose graph initialization over existing local methods.

Findings

Outperforms state-of-the-art retrieval methods on ambiguous scenes.

Produces more reliable and compact pose graphs.

Enhances reconstruction accuracy in sparse and high-speed scenarios.

Abstract

The pose graph is a core component of Structure-from-Motion (SfM), where images act as nodes and edges encode relative poses. Since geometric verification is expensive, SfM pipelines restrict the pose graph to a sparse set of candidate edges, making initialization critical. Existing methods rely on image retrieval to connect each image to its $k$ nearest neighbors, treating pairs independently and ignoring global consistency. We address this limitation through the concept of edge prioritization, ranking candidate edges by their utility for SfM. Our approach has three components: (1) a GNN trained with SfM-derived supervision to predict globally consistent edge reliability; (2) multi-minimal-spanning-tree-based pose graph construction guided by these ranks; and (3) connectivity-aware score modulation that reinforces weak regions and reduces graph diameter. This globally informed initialization yields more reliable and compact pose graphs, improving reconstruction accuracy in sparse and high-speed settings and outperforming SOTA retrieval methods on ambiguous scenes. The ode and trained models are available at https://github.com/weitong8591/global_edge_prior.