Research map / 3D vision / Human pose estimation
NeFSAC: Neurally Filtered Minimal Samples
Luca Cavalli, Marc Pollefeys, Daniel Barath
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 NeFSAC neural filtering | USAC · KITTI, essential matrix | run-time (ms): 234.8 -> 69.7 | Improved |
| add NeFSAC neural filtering | USAC · KITTI, fundamental matrix | run-time (ms): 213.4 -> 85.9 | Improved |
| add NeFSAC neural filtering | USAC · Malaga, essential matrix | run-time (ms): 350.0 -> 34.0 | Improved |
| add NeFSAC neural filtering | USAC · Malaga, essential matrix | rotation error (deg): 3.3 -> 1.9 | Improved |
| add NeFSAC neural filtering | USAC · KITTI, essential matrix | rotation error (deg): 4.3 -> 4.3 | No clear effect |
| add NeFSAC neural filtering | USAC · PhotoTourism, essential matrix | run-time (ms): 805.1 -> 76.5 | Improved |
| add NeFSAC neural filtering | USAC · PhotoTourism, fundamental matrix | run-time (ms): 154.6 -> 61.7 | Improved |
| add NeFSAC neural filtering | USAC · PhotoTourism, essential matrix | translation error (deg): 7.9 -> 6.1 | Improved |
| add NeFSAC neural filtering | MAGSAC++ · Malaga, essential matrix | run-time (ms): 1456 -> 122 | Improved |
| add NeFSAC neural filtering | MAGSAC++ · PhotoTourism British Museum | AUC@10: 0.82 -> 0.83 | Improved |
| train on KITTI instead of PhotoTourism (domain shift) (ablation) | NeFSAC · PhotoTourism, essential matrix | run-time (ms): 76.5 -> 103.0 | Mixed |
| add NeFSAC to USAC with oracle inlier predictor | USAC + Oracle · KITTI | AUC@10: 0.93 -> 0.95 | Improved |
Closest papers by meaning
- On the Instability of Relative Pose Estimation and RANSAC's Role (CVPR 2022 · similarity 0.86)
- Learning To Find Good Models in RANSAC (CVPR 2022 · similarity 0.85)
- RANRAC: Robust Neural Scene Representations via Random Ray Consensus (ECCV 2024 · similarity 0.85)
- L-SWAG: Layer-Sample Wise Activation with Gradients Information for Zero-Shot NAS on Vision Transformers (CVPR 2025 · similarity 0.83)
- ICON: Incremental CONfidence for Joint Pose and Radiance Field Optimization (CVPR 2024 · similarity 0.83)
- Learning Neural Exposure Fields for View Synthesis (NeurIPS 2025 · similarity 0.83)
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.
