Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.21963 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915816216199168 |
|---|---|
| author | Wei, Tong Tolias, Giorgos Matas, Jiri Barath, Daniel |
| author_facet | Wei, Tong Tolias, Giorgos Matas, Jiri Barath, Daniel |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_21963 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Global-Aware Edge Prioritization for Pose Graph Initialization Wei, Tong Tolias, Giorgos Matas, Jiri Barath, Daniel Computer Vision and Pattern Recognition 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. |
| title | Global-Aware Edge Prioritization for Pose Graph Initialization |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.21963 |