Research map / Point clouds / Shape and surface completion
Ex-Sim(3)-Reg: 2D-3D Correspondence Pruning via Extended Sim(3) Registration
Pei An, Muyao Peng, Junfeng Ding, Jiaqi Yang, Liangliang Nan
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 scale-correction function to FastMAC pruning (ablation) | Ex-Sim(3)-Reg · 7-Scenes (noisy) | Registration Recall: 0.311 -> 0.472 | Improved |
| use full MAC instead of FastMAC search (ablation) | Ex-Sim(3)-Reg · 7-Scenes (noisy) | Registration Recall: 0.472 -> 0.485; 73.6 -> 204.8 ms | Mixed |
| increase number of sub-intervals K from 4 to 8 (ablation) | Ex-Sim(3)-Reg · 7-Scenes (noisy) | Registration Recall: 0.362 -> 0.472 | Improved |
| increase K from 8 to 12 (ablation) | Ex-Sim(3)-Reg · 7-Scenes (noisy) | Registration Recall: 0.472 -> 0.427 | Hurt |
| raise pixel reprojection threshold from 12 to 16 (ablation) | Ex-Sim(3)-Reg · 7-Scenes (noisy) | Registration Recall: 0.466 -> 0.472 | Improved |
| raise pixel reprojection threshold from 16 to 24 (ablation) | Ex-Sim(3)-Reg · 7-Scenes (noisy) | Registration Recall: 0.472 -> 0.372 | Hurt |
| lower pixel reprojection threshold from 16 to 8 (ablation) | Ex-Sim(3)-Reg · 7-Scenes (noisy) | Registration Recall: 0.472 -> 0.380 | Hurt |
| replace regular Sim(3) pruning (TurboReg) with extended Sim(3) pruning | Sim(3)-TurboReg · 7-Scenes (disturbed) | Registration Recall: 0.343 -> 0.472 | Improved |
| replace regular Sim(3) pruning with Ex-Sim(3)-Reg under low-inlier condition | Sim(3)-TurboReg · RGBD-v2 (low-inlier) | Registration Recall: 0.344 -> 0.542 | Improved |
| regular Sim(3) pruning instead of default RANSAC baseline | Sim(3)-SC2-PCR · ScanNet (7-Scenes to ScanNet) | Registration Recall: 0.431 -> 0.260 | Hurt |
| add Ex-Sim(3)-Reg to Matr | Matr · ScanNet (cross-domain I2P) | Registration Recall: 0.842 -> 0.943 | Improved |
| add Ex-Sim(3)-Reg to FreeReg | FreeReg · ScanNet (cross-domain I2P) | Registration Recall: 0.780 -> 0.932 | Improved |
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.
- FreeReg — FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimators (ICLR 2024) · 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
- FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimators (ICLR 2024 · similarity 0.86)
- Generative Point Cloud Registration (ICML 2025 · similarity 0.86)
- Scalable 3D Registration via Truncated Entry-wise Absolute Residuals (CVPR 2024 · similarity 0.86)
- Adaptive 3D Reconstruction via Diffusion Priors and Forward Curvature-Matching Likelihood Updates (NeurIPS 2025 · similarity 0.86)
- PF3plat: Pose-Free Feed-Forward 3D Gaussian Splatting for Novel View Synthesis (ICML 2025 · similarity 0.85)
- Progressive Correspondence Regenerator for Robust 3D Registration (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.
