Research map / 3D vision / Depth and stereo
Learning Dense Flow Field for Highly-accurate Cross-view Camera Localization
Zhenbo Song, Xianghui Ze, Jianfeng Lu, Yujiao Shi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Ground Feature Projection (ablation) | dense flow cross-view localizer · KITTI same-area | location error mean (m): 3.52 -> 1.48 | Improved |
| add BEV Feature RefineBlock (ablation) | dense flow cross-view localizer · KITTI same-area | location error mean (m): 2.02 -> 1.48 | Improved |
| add Matching Loss (ablation) | dense flow cross-view localizer · KITTI same-area | location error mean (m): 9.06 -> 1.48 | Improved |
| add Position Loss (ablation) | dense flow cross-view localizer · KITTI same-area | location error mean (m): 0.99 -> 1.48 | Mixed |
| add Position Loss (ablation) | dense flow cross-view localizer · KITTI cross-area | location error mean (m): 9.63 -> 7.97 | Improved |
| add Confidence mechanism (ablation) | dense flow cross-view localizer · KITTI same-area | location error mean (m): 1.65 -> 1.48 | Improved |
| add Confidence mechanism (ablation) | dense flow cross-view localizer · KITTI cross-area | location error mean (m): 7.76 -> 7.97 | No clear effect |
| replace SliceMatch with dense-flow method | cross-view localization · KITTI same-area | location error mean (m): 7.96 -> 1.48 | Improved |
| replace SliceMatch with dense-flow method | cross-view localization · KITTI cross-area | location error mean (m): 13.50 -> 7.97 | Improved |
| replace SliceMatch with dense-flow method | cross-view localization · VIGOR same-area aligned | location error mean (m): 5.18 -> 3.03 | Improved |
| replace SliceMatch with dense-flow method | cross-view localization · VIGOR cross-area aligned | location error mean (m): 5.53 -> 5.01 | Improved |
| replace SliceMatch with dense-flow method, unknown orientation | cross-view localization · VIGOR same-area unaligned | azimuth error mean (deg): 28.43 -> 11.20 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SliceMatch — SliceMatch: Geometry-Guided Aggregation for Cross-View Pose Estimation (CVPR 2023) · better on 13 of 13 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Boosting 3-DoF Ground-to-Satellite Camera Localization Accuracy via Geometry-Guided Cross-View Transformer (ICCV 2023 · similarity 0.89)
- BevSplat: Resolving Height Ambiguity via Feature-Based Gaussian Primitives for Weakly-Supervised Cross-View Localization (NeurIPS 2025 · similarity 0.89)
- Seeing Across Skies and Streets: Feedforward 3D Reconstruction from Satellite, Drone, and Ground Images (NeurIPS 2026 · similarity 0.87)
- Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations (CVPR 2025 · similarity 0.87)
- FG^2: Fine-Grained Cross-View Localization by Fine-Grained Feature Matching (CVPR 2025 · similarity 0.87)
- View From Above: Orthogonal-View aware Cross-view Localization (CVPR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
