Research map / 3D vision / Depth and stereo
AtLoc: Attention Guided Camera Localization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 self-attention to the single-image pose regression model (ablation) | AtLoc on 7 Scenes · 7 Scenes | average median position error: 0.22m -> 0.20m | Helped |
| Add self-attention to the single-image pose regression model (ablation) | AtLoc on Oxford RobotCar · Oxford RobotCar | average mean position error: 77.70m -> 23.8m | Helped |
| Replace the attention module with an LSTM module (ablation) | AtLoc on 7 Scenes · 7 Scenes | average median position error: 0.20m -> 0.21m | Hurt |
| Replace the attention module with an LSTM module (ablation) | AtLoc on Oxford RobotCar · Oxford RobotCar | average mean position error: 23.8m -> 44.60m | Hurt |
| Add temporal constraints to AtLoc (ablation) | AtLoc on 7 Scenes · 7 Scenes | average median position error: 0.20m -> 0.19m | Helped |
| Add temporal constraints to AtLoc (ablation) | AtLoc on Oxford RobotCar · Oxford RobotCar | average mean position error: 23.8m -> 19.7m | Helped |
| Use single-image self-attention rather than sequential MapNet | AtLoc on Oxford RobotCar · Oxford RobotCar LOOP and FULL | average median position error: 12.17m -> 8.54m | Helped |
| Use attention-guided features instead of PoseNet+ features | AtLoc on Oxford RobotCar · Oxford RobotCar LOOP and FULL | average median position error: 55.52m -> 8.54m | Helped |
Closest papers by meaning
- Learning To Detect Scene Landmarks for Camera Localization (CVPR 2022 · similarity 0.89)
- Reloc3r: Large-Scale Training of Relative Camera Pose Regression for Generalizable, Fast, and Accurate Visual Localization (CVPR 2025 · similarity 0.88)
- Back to the Feature: Learning Robust Camera Localization From Pixels To Pose (CVPR 2021 · similarity 0.88)
- Geometry-Aware Learning of Maps for Camera Localization (CVPR 2018 · similarity 0.87)
- HyperPose: Hypernetwork-Infused Camera Pose Localization and an Extended Cambridge Landmarks Dataset (CVPR 2025 · similarity 0.85)
- LightLoc: Learning Outdoor LiDAR Localization at Light Speed (CVPR 2025 · similarity 0.84)
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.