AI Paper Atlas

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

CVPR 2026 · 3D vision › Depth and stereo · approach: Scene coordinate regression for localization · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
use 120 deg FoV instead of 90 deg FoV training imagesOrienterNet · CV-RHO cleanPR@5m: 71.04 -> 78.03Improved
use 360 deg panoramas directly in OrienterNetOrienterNet · CV-RHO cleanPR@5m: 78.03 -> 28.78Hurt
replace OrienterNet with RHOOrienterNet 120 deg · CV-RHO cleanPR@5m: 78.03 -> 84.36Improved
add POF fusion over no fusion (ablation)RHO · Washington cleanOR@5deg: 69.87 -> 91.21Improved
add POF fusion over no fusion (ablation)RHO · Washington cleanPR@5m: 80.60 -> 83.16Improved
use mutual POF instead of position-prior-only fusion (ablation)RHO · Washington cleanPR@5m: 77.38 -> 83.16Improved
use mutual POF instead of orientation-prior-only fusion (ablation)RHO · Washington cleanOR@5deg: 88.69 -> 91.21Improved
split panorama into three 120 deg views instead of four 90 deg views (ablation)RHO SUM · Washington cleanPR@5m: 34.22 -> 80.60Improved
split panorama into three 120 deg views instead of 60 deg views (ablation)RHO SUM · Washington cleanPR@5m: 7.34 -> 80.60Improved
train RHO on each noisy variation instead of clean onlyRHO · CV-RHO motion blurPR@5m: 49.69 -> 84.07Improved
train on each noisy variation instead of clean onlyOrienterNet · CV-RHO average degradation PR@3mPR@3m: -17.41 -> -3.45Improved
evaluate RHO cross-region instead of same-regionRHO · Mount Vernon cleanPR@5m: 84.36 -> 73.79Hurt

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.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy