Research map / 3D vision / Depth and stereo
Scene Coordinate Reconstruction: Posing of Image Collections via Incremental Learning of a Relocalizer
Eric Brachmann, Jamie Wynn, Shuai Chen, Tommaso Cavallari, Aron Monszpart, Daniyar Turmukhambetov, Victor Adrian Prisacariu
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 |
|---|---|---|---|
| initialize ACE0 with KinectFusion poses and refine (KF+ACE0) | ACE0 · 7-Scenes (all frames) | PSNR (dB): 18.7 -> 21.5 | Improved |
| use ACE0 self-supervised scene coordinate reconstruction instead of COLMAP default | 7-Scenes poses · 7-Scenes (all frames) | PSNR (dB): 21.2 -> 21.2 | No clear effect |
| ACE0 versus COLMAP on Mip-NeRF 360 | Mip-NeRF 360 poses · Mip-NeRF 360 | PSNR (dB): 24.7 -> 22.9 | Hurt |
| ACE0 versus NoPe-NeRF and BARF on 200-frame 7-Scenes | learning-based SfM · 7-Scenes (200 frames) | PSNR (dB): 12.6 -> 21.0 | Improved |
| ACE0 versus DUSt3R on 50-frame 7-Scenes | DUSt3R · 7-Scenes (50 frames) | PSNR (dB): 17.1 -> 17.9 | Improved |
| train ACE self-supervised via ACE0 instead of with COLMAP poses | ACE relocalizer · 7-Scenes relocalization | % poses under 5cm, 5deg: 97.6 -> 93.8 | Hurt |
| train ACE with KinectFusion poses instead of ACE0 poses, evaluated against COLMAP | ACE relocalizer · 7-Scenes relocalization | % poses under 5cm, 5deg: 93.8 -> 71.5 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- NoPe-NeRF — NoPe-NeRF: Optimising Neural Radiance Field With No Pose Prior (CVPR 2023) · better on 2 of 2 shared results
- BARF — BARF: Bundle-Adjusting Neural Radiance Fields (ICCV 2021) · better on 2 of 2 shared results
- DUSt3R — DUSt3R: Geometric 3D Vision Made Easy (CVPR 2024) · better on 1 of 1 shared result
- DROID-SLAM — DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras (NeurIPS 2021) · better on 1 of 1 shared result
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
- Accelerated Coordinate Encoding: Learning to Relocalize in Minutes Using RGB and Poses (CVPR 2023 · similarity 0.93)
- ACE-G: Improving Generalization of Scene Coordinate Regression Through Query Pre-Training (ICCV 2025 · similarity 0.90)
- Reloc3r: Large-Scale Training of Relative Camera Pose Regression for Generalizable, Fast, and Accurate Visual Localization (CVPR 2025 · similarity 0.89)
- Map-Free Visual Relocalization: Metric Pose Relative to a Single Image (ECCV 2022 · similarity 0.89)
- Hierarchical Visual Relocalization with Nearest View Synthesis from Feature Gaussian Splatting (CVPR 2026 · similarity 0.89)
- Scene Coordinate Reconstruction Priors (ICCV 2025 · similarity 0.89)
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.
