AI Paper Atlas

Research map / 3D vision / Depth and stereo

HOLO: Homography-Guided Pose Estimator Network for Fine-Grained Visual Localization on SD Maps

Xuchang Zhong, Xu Cao, Jinke Feng, Hao Fang

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
replace 3-DoF pose regression head with homography head (ablation)HOLO-CA · nuScenes valPosition Recall@1m: 10.06 -> 21.47Improved
replace 3-DoF pose regression head with homography head (ablation)HOLO (one iteration) · nuScenes valPosition Recall@1m: 9.88 -> 26.57Improved
homography head on orientation (ablation)HOLO (one iteration) · nuScenes valAOE (deg): 2.19 -> 1.49Improved
increase refinement iterations from 1 to 6 (ablation)HOLO · nuScenes valPosition Recall@1m: 26.57 -> 36.41Improved
increase iterations from 2 to 6 on coarse metrics (ablation)HOLO · nuScenes valOrientation Recall@5deg: 98.14 -> 97.35No clear effect
homography-guided fusion instead of cross-attention fusion (ablation)HOLO · nuScenes valPosition Recall@1m: 21.47 -> 26.57Improved
add joint homography learning to BEV perception training (ablation)BEV segmentation · nuScenes valIoU drivable area: 76.52 -> 79.89Improved
add joint homography learning to BEV perception training (ablation)BEV segmentation · nuScenes valIoU buildings: 68.41 -> 69.43Improved
use coarser BEV grid 0.5 mpp instead of 0.25 (ablation)HOLO · nuScenes valPosition Recall@1m: 36.41 -> 36.10No clear effect
increase simulated map noise from 30 m to 40 m (ablation)HOLO · nuScenes valPosition Recall@1m: 36.41 -> 18.24Hurt
supervise with SD road layer instead of HD drivable layer (ablation)HOLO · nuScenes valPosition Recall@1m: 36.41 -> 30.40Hurt

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