Research map / 3D vision / Human pose estimation
NOPE-SAC: Neural One-Plane RANSAC for Sparse-View Planar 3D Reconstruction
Bin Tan, Nan Xue, Tianfu Wu, Gui-Song Xia
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 AGNN to plane matching (ablation) | NOPE-SAC plane matching · Matterport3D | F-score: 48.8 -> 50.7 | Improved |
| add geometric affinity S_g (ablation) | NOPE-SAC plane matching · Matterport3D | F-score: 42.6 -> 50.7 | Improved |
| add plane warping before correspondence embedding (ablation) | NOPE-SAC-Reg · Matterport3D | translation accuracy within 0.5m: 42.3% -> 48.3% | Improved |
| use Soft fusion instead of Avg (ablation) | NOPE-SAC-Reg · Matterport3D | rotation accuracy within 15 deg: 77.1% -> 86.9% | Improved |
| use Soft fusion instead of Min-Cost (ablation) | NOPE-SAC-Reg · Matterport3D | translation accuracy within 0.5m: 45.4% -> 48.3% | Improved |
| use Soft fusion instead of Max-Score (ablation) | NOPE-SAC-Reg · Matterport3D | translation accuracy within 0.5m: 43.1% -> 48.3% | Improved |
| add NOPE-SAC refinement to initial pose | NOPE-SAC-Reg · Matterport3D | rotation median error (deg): 11.16 -> 2.77 | Improved |
| replace PlaneFormers-TR refinement with NOPE-SAC (ablation) | regressive initial pose · Matterport3D | rotation accuracy within 10 deg: 53.5% -> 84.0% | Improved |
| use regressive initial pose instead of classification pose | NOPE-SAC · Matterport3D | rotation accuracy within 10 deg: 80.2% -> 84.0% | Improved |
| add AIM on top of warping (ablation) | NOPE-SAC-Reg · Matterport3D | translation accuracy within 0.5m: 49.3% -> 48.3% | No clear effect |
| refine SuperGlue pose with NOPE-SAC | SuperGlue · Matterport3D | rotation accuracy within 15 deg: 71.0% -> 82.2% | Improved |
| refine LoFTR pose with NOPE-SAC | LoFTR · ScanNet | translation accuracy within 0.5m: 59.5% -> 59.1% | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SuperGlue — SuperGlue: Learning Feature Matching With Graph Neural Networks (CVPR 2020) · better on 2 of 2 shared results
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
- Sparse-view Pose Estimation and Reconstruction via Analysis by Generative Synthesis (NeurIPS 2024 · similarity 0.88)
- Towards In-the-wild 3D Plane Reconstruction from a Single Image (CVPR 2025 · similarity 0.88)
- LEAP: Liberate Sparse-View 3D Modeling from Camera Poses (ICLR 2024 · similarity 0.87)
- No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Images (ICLR 2025 · similarity 0.86)
- Sparse3D: Distilling Multiview-Consistent Diffusion for Object Reconstruction from Sparse Views (AAAI 2024 · similarity 0.85)
- Pow3R: Empowering Unconstrained 3D Reconstruction with Camera and Scene Priors (CVPR 2025 · similarity 0.85)
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.
