Research map / 3D vision / Depth and stereo
RHO: Robust Holistic OSM-Based Metric Cross-View Geo-Localization
Junwei Zheng, Ruize Dai, Ruiping Liu, Zichao Zeng, Yufan Chen, Fangjinhua Wang, Kunyu Peng, Kailun Yang, Jiaming Zhang, Rainer Stiefelhagen
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 |
|---|---|---|---|
| use 120 deg FoV instead of 90 deg FoV training images | OrienterNet · CV-RHO clean | PR@5m: 71.04 -> 78.03 | Improved |
| use 360 deg panoramas directly in OrienterNet | OrienterNet · CV-RHO clean | PR@5m: 78.03 -> 28.78 | Hurt |
| replace OrienterNet with RHO | OrienterNet 120 deg · CV-RHO clean | PR@5m: 78.03 -> 84.36 | Improved |
| add POF fusion over no fusion (ablation) | RHO · Washington clean | OR@5deg: 69.87 -> 91.21 | Improved |
| add POF fusion over no fusion (ablation) | RHO · Washington clean | PR@5m: 80.60 -> 83.16 | Improved |
| use mutual POF instead of position-prior-only fusion (ablation) | RHO · Washington clean | PR@5m: 77.38 -> 83.16 | Improved |
| use mutual POF instead of orientation-prior-only fusion (ablation) | RHO · Washington clean | OR@5deg: 88.69 -> 91.21 | Improved |
| split panorama into three 120 deg views instead of four 90 deg views (ablation) | RHO SUM · Washington clean | PR@5m: 34.22 -> 80.60 | Improved |
| split panorama into three 120 deg views instead of 60 deg views (ablation) | RHO SUM · Washington clean | PR@5m: 7.34 -> 80.60 | Improved |
| train RHO on each noisy variation instead of clean only | RHO · CV-RHO motion blur | PR@5m: 49.69 -> 84.07 | Improved |
| train on each noisy variation instead of clean only | OrienterNet · CV-RHO average degradation PR@3m | PR@3m: -17.41 -> -3.45 | Improved |
| evaluate RHO cross-region instead of same-region | RHO · Mount Vernon clean | PR@5m: 84.36 -> 73.79 | Hurt |
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.
- OrienterNet — OrienterNet: Visual Localization in 2D Public Maps With Neural Matching (CVPR 2023) · better on 3 of 3 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
- Sector-Level Cross-View Geo-Localization with Implicit Orientation via Azimuthal Scanning (ECCV 2026 · similarity 0.88)
- SinGeo: Unlock Single Model's Potential for Robust Cross-View Geo-Localization (CVPR 2026 · similarity 0.87)
- Aligning Geometric Spatial Layout in Cross-View Geo-Localization via Feature Recombination (AAAI 2024 · similarity 0.86)
- CVGL: Causal Learning and Geometric Topology (NeurIPS 2025 · similarity 0.85)
- Unleashing Unlabeled Data: A Paradigm for Cross-View Geo-Localization (CVPR 2024 · similarity 0.85)
- From Coarse to Fine: A Matching and Alignment Framework for Unsupervised Cross-View Geo-Localization (AAAI 2025 · similarity 0.85)
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.
