Search papers, labs, and topics across Lattice.
The paper introduces a novel approach to pose graph initialization for Structure-from-Motion (SfM) by prioritizing edges based on global consistency. They train a Graph Neural Network (GNN) with SfM supervision to predict edge reliability, construct pose graphs using multi-minimal-spanning-trees guided by these ranks, and apply connectivity-aware score modulation. The resulting globally-informed initialization produces more reliable and compact pose graphs, leading to improved reconstruction accuracy, especially in challenging scenarios.
By learning to prioritize edges based on global consistency, this method constructs more reliable and compact pose graphs for Structure-from-Motion, outperforming traditional nearest neighbor retrieval, especially in ambiguous scenes.
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.
