| Paper | Code | AUC@5 | AUC@10 | AUC@20 | ModelName | ReleaseDate |
|---|---|---|---|---|---|---|
| Marginalized Bundle Adjustment: Multi-View Camera Pose from Monocular Depth Estimates | ✓ Link | 32.4 | 54.1 | 71.0 | RoMa + MBA | 2026-02-21 |
| RoMa: Robust Dense Feature Matching | ✓ Link | 31.8 | 53.4 | 70.9 | RoMa | 2023-05-24 |
| Multi-Session SLAM with Differentiable Wide-Baseline Pose Optimization | ✓ Link | 30.5 | 50.9 | 67.5 | MultiSlam-DiffPose | 2024-04-23 |
| LoFTR: Detector-Free Local Feature Matching with Transformers | ✓ Link | 22.06 | 40.8 | LoFTR-DS | 2021-04-01 | |
| XFeat Revisited: Reproducibility and Evaluation of a Lightweight Image Matcher | ✓ Link | 19.9 | 37.1 | 53.2 | XFeat* Reproduced | 2026-08-10 |
| Fast Relative Pose Estimation using Relative Depth | 18.39 | 35.46 | 52.24 | 3 pt. + RelScaleNet | 2024-03-18 | |
| SuperGlue: Learning Feature Matching with Graph Neural Networks | ✓ Link | 16.16 | 33.81 | 51.84 | SuperPoint + SuperGlue | 2019-11-26 |
| VGGT: Visual Geometry Grounded Transformer | ✓ Link | 55.2 | 73.4 | VGGT | 2025-03-14 | |
| DINOv3 | ✓ Link | 56.1 | DINOv3 + VGGT | 2025-08-13 |