Overview

Pose graph optimization represents robot or camera poses as graph nodes and odometry, loop closures, or landmark relations as edges, then adjusts the poses to reduce disagreement across measurements. It is a core tool in SLAM, mapping, and reconstruction, with nonlinear least squares on SE(2), SE(3), or Sim(3) commonly paired with robust methods such as switchable constraints to handle outliers. Current approaches include distributed and Riemannian solvers for multi-robot settings, learned edge selection and adaptive optimization, and hierarchical or certifiable methods for faster or more reliable convergence. A central tension is balancing accuracy and robustness against computation and communication costs, while pose-only constraints can miss scale, uncertainty, or map structure needed for consistent reconstruction.

Latest papers

31 papers2 papers in the last four weeks, down 71% on the four weeks before. 0.0% of all new papers.

Feb 25, 2026

Global-Aware Edge Prioritization for Pose Graph Initialization

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 kk 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.