AI Paper Atlas

Research map / 3D vision / Depth and stereo

AutoCompass: Accurate Visual Localization on Public Maps by Learning from Weak Labels

Javier Tirado-Garín, Alan Savio Paul, Shuai Chen, Axel Barroso-Laguna, Tommaso Cavallari, Daniyar Turmukhambetov, Victor Adrian Prisacariu, Eric Brachmann

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
train with raw GPS and chunk marginalization instead of strong geo-referenced 3-DoF labelsOrienterNet (ResNet-101) · KITTI (no orientation prior)lateral recall@1m: 37.7 -> 43.1Improved
add relative-pose supervision instead of raw GPS onlyAutoCompass (ResNet-101) · KITTI (no orientation prior)lateral recall@1m: 43.1 -> 56.6Improved
swap ResNet-101 for DINOv2-B backboneAutoCompass w/ rel. poses · KITTI (no orientation prior)lateral recall@1m: 56.6 -> 70.8Improved
add relative-pose supervision instead of raw GPS onlyAutoCompass (ResNet-101) · LaMAriaXY recall@5m: 26.3 -> 33.0Improved
add relative-pose supervision instead of raw GPS onlyAutoCompass (ResNet-101) · Oxford Day-and-NightXY recall@5m: 52.2 -> 56.4Improved
train with raw GPS instead of strong supervision on ODNOrienterNet (ResNet-101) · Oxford Day-and-NightXY recall@1m: 9.8 -> 10.4No clear effect
add GPS neighborhood marginalization (r=5 m vs r=0) (ablation)AutoCompass GPS · Average over benchmarksXY recall@5m: 37.1 -> 44.4Improved
increase chunk radius from 20 m to 30 m (ablation)AutoCompass GPS · Average over benchmarksXY recall@5m: 45.0 -> 37.6Hurt
vary norm binning size from 1 to 10 m (ablation)AutoCompass dTheta+norm · Average over benchmarksXY recall@5m: 46.5 -> 44.9No clear effect
remove GPS chunk loss from combined supervision (ablation)AutoCompass · Average over benchmarksXY recall@5m: 48.1 -> 50.8Improved
fuse single-view estimates over under 10 framesAutoCompass · LaMAria (sequential)XY recall@5m: 33.0 -> 92.6Improved
add relative rotation loss to chunk GPS (ablation)AutoCompass (ResNet-101) · LaMAriaorientation recall@5deg: 28.7 -> 31.9Improved

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