AI Paper Atlas

Research map / 3D vision / Depth and stereo

Warp-free Cross-view Geo-localization via Feature-space Consensus Mining

Zhuo Song, Lian Xu, Runqing Jiang, Kunhong Li, Yongjian Zhang, Ye Zhang, Yulan Guo

ECCV 2026 · 3D vision › Depth and stereo · approach: Scene coordinate regression for localization · 0 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
replace average pooling with separate attention probes (SPP) (ablation)Sample4Geo-style baseline · VIGOR same-areaR@1 (%): 77.86 -> 79.96Improved
replace SPP with global shared probes (GPP) (ablation)probe pooling without CM · VIGOR same-areaR@1 (%): 79.96 -> 81.59Improved
add joint-view consensus mediator (CM) to SPP (ablation)SPP · VIGOR same-areaR@1 (%): 79.96 -> 81.96Improved
add joint-view consensus mediator (CM) to GPP (ablation)GPP · VIGOR same-areaR@1 (%): 81.59 -> 83.15Improved
increase probe count from 16 to 64 (ablation)GPP · VIGOR same-areaR@1 (%): 82.24 -> 83.15Improved
increase probe count from 64 to 96 (ablation)GPP · VIGOR same-areaR@1 (%): 83.15 -> 83.15No clear effect
use consensus mining instead of Co-RetrievalOurs · CVUSAR@1 (%): 98.71 -> 99.29Improved
use consensus mining instead of Co-RetrievalOurs · VIGOR cross-areaR@1 (%): 72.19 -> 73.60Improved
use consensus mining instead of Co-RetrievalOurs · VIGOR same-areaR@1% (%): 99.70 -> 99.59Hurt
use consensus mining instead of QDFLOurs · University-1652 Drone->SatelliteR@1 (%): 95.00 -> 97.42Improved

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