Research map / 3D vision / Depth and stereo
A Scene is Worth a Thousand Features: Feed-Forward Camera Localization from a Collection of Image Features
Axel Barroso-Laguna, Tommaso Cavallari, Victor Adrian Prisacariu, Eric Brachmann
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 |
|---|---|---|---|
| add scale normalization of mapping translations (ablation) | FastForward · Cambridge Landmarks | accuracy 10cm 10deg (%): 1.8 -> 26.7 | Improved |
| add scale normalization of mapping translations (ablation) | FastForward · Wayspots | accuracy 10cm 10deg (%): 47.0 -> 51.4 | Improved |
| add scale normalization of mapping translations (ablation) | FastForward · Indoor6 | accuracy 10cm 10deg (%): 83.4 -> 91.5 | Improved |
| add scale normalization of mapping translations (ablation) | FastForward · RIO10 (val) | accuracy 10cm 10deg (%): 35.9 -> 40.6 | Improved |
| add scale normalization of mapping translations (ablation) | FastForward · 7-Scenes | accuracy 10cm 10deg (%): 89.1 -> 90.2 | Improved |
| select mapping images by retrieval instead of random sampling (ablation) | FastForward · Wayspots | accuracy 10cm 10deg (%): 43.9 -> 51.4 | Improved |
| select mapping images by retrieval instead of uniform sampling (ablation) | FastForward · Wayspots | accuracy 10cm 10deg (%): 47.8 -> 51.4 | Improved |
| remove retrieval step | Reloc3r · Wayspots | accuracy 10cm 10deg (%): 37.1 -> 19.7 | Hurt |
| increase number of mapping images in map representation (ablation) | FastForward · Map-free val | accuracy (all thresholds) | Improved |
| increase number of mapping features from 150 to 3,000 (ablation) | FastForward · Map-free val | accuracy 10cm 10deg / 25cm 5deg | No clear effect |
| increase number of mapping features (ablation) | FastForward · Map-free val | accuracy 10cm 1deg | Improved |
| use scene-specific network training (ACE) instead of FastForward on Wayspots translation | FastForward · Wayspots | median translation error (m): 1.33 -> 0.17 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- GLACE — GLACE: Global Local Accelerated Coordinate Encoding (CVPR 2024) · better on 2 of 4 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
- Investigating the Role of Image Retrieval for Visual Localization (IJCV 2022 · similarity 0.88)
- Geometry-Aware Learning of Maps for Camera Localization (CVPR 2018 · similarity 0.87)
- WildPose: A Unified Framework for Robust Pose Estimation in the Wild (CVPR 2026 · similarity 0.87)
- Photo-SLAM: Real-time Simultaneous Localization and Photorealistic Mapping for Monocular Stereo and RGB-D Cameras (CVPR 2024 · similarity 0.86)
- Map-Relative Pose Regression for Visual Re-Localization (CVPR 2024 · similarity 0.86)
- RELOCATE: A Simple Training-Free Baseline for Visual Query Localization Using Region-Based Representations (CVPR 2025 · similarity 0.86)
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.
