Research map / Video / Video segmentation and tracking
VidLoc: A Deep Spatio-Temporal Model for 6-DoF Video-Clip Relocalization
Check 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 |
|---|---|---|---|
| Use a recurrent video-clip model instead of independent single-frame pose prediction | VidLoc · 7-Scenes | localization accuracy: 23.4% - 55% increase in accuracy | Helped |
| Use VidLoc (RGB) instead of Posenet | 7-Scenes Office · 7-Scenes Office | translation error: 0.48m -> 0.26m | Helped |
| Use VidLoc (RGB) instead of Posenet | 7-Scenes Stairs · 7-Scenes Stairs | translation error: 0.47m -> 0.26m | Helped |
| Use VidLoc (RGB) instead of Score-Forest | 7-Scenes · 7-Scenes | translation error | Mixed |
| Increase the video sequence/window length | VidLoc · 7-Scenes | localization error | Helped |
| Increase the video sequence length | VidLoc · Oxford RobotCar | localization accuracy | Helped |
| Use sequence length 50 or 100 rather than a single image | VidLoc · Oxford RobotCar | trajectory localization: 50 and 100 | Helped |
| Use VidLoc instead of Posenet | Oxford RobotCar, sequence length 100 · Oxford RobotCar | poses within 20 meters: more than half; less than 15% | Helped |
| Use RGB-D input instead of RGB-only input | VidLoc · 7-Scenes | translation error | Helped |
| Use VidLoc with RGB-D rather than Score-Forest | 7-Scenes · 7-Scenes | translation error | Hurt |
Closest papers by meaning
- Learning Less Is More - 6D Camera Localization via 3D Surface Regression (CVPR 2018 · similarity 0.87)
- From Coarse to Fine: Robust Hierarchical Localization at Large Scale (CVPR 2019 · similarity 0.86)
- Reloc3r: Large-Scale Training of Relative Camera Pose Regression for Generalizable, Fast, and Accurate Visual Localization (CVPR 2025 · similarity 0.86)
- From Sparse to Dense: Camera Relocalization with Scene-Specific Detector from Feature Gaussian Splatting (CVPR 2025 · similarity 0.85)
- Transferable Video Moment Localization by Moment-Guided Query Prompting (AAAI 2024 · similarity 0.84)
- RELOCATE: A Simple Training-Free Baseline for Visual Query Localization Using Region-Based Representations (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.