Research map / Point clouds / Registration and matching
Non-rigid Point Cloud Registration with Neural Deformation Pyramid
YANG LI, Tatsuya Harada
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 |
|---|---|---|---|
| use hierarchical pyramid of small MLPs instead of single full-band coordinate MLP (Nerfies) | NDP · 4DMatch | Time (s): 115.94 -> 2.31 | Improved |
| pyramid decomposition reduces gradient iterations | NDP · 4DMatch | iterations to converge: 3792 -> 738 | Improved |
| use L1 instead of L2 Chamfer distance (ablation) | NDP · 4DMatch | AccS (%): 14.34 -> 18.69 | Improved |
| use L1 instead of L2 correspondence term (ablation) | LNDP · 4DLoMatch | AccS (%): 26.59 -> 28.65 | Improved |
| use SE(3) warp field instead of R3 vector field (ablation) | LNDP · 4DMatch | AccS (%): 58.03 -> 62.85 | Improved |
| use SE(3) instead of Sim(3) warp field (ablation) | LNDP · 4DMatch | AccS (%): 59.33 -> 62.85 | Improved |
| axis-angle rotation instead of quaternion (ablation) | LNDP · 4DMatch | AccS (%): 55.47 -> 62.85 | Improved |
| axis-angle vs Euler angle rotation (ablation) | LNDP · 4DMatch | AccS (%): 62.72 -> 62.85 | No clear effect |
| add Transformer outlier rejection of correspondences (ablation) | LNDP · 4DMatch | AccS (%): 54.89 -> 61.27 | Improved |
| xavier_uniform instead of ones_ init (ablation) | NDP · 4DMatch | AccS (%): 0.00 -> 14.34 | Improved |
| kaiming_uniform vs xavier_uniform init (ablation) | NDP · 4DMatch | AccS (%): 14.93 vs 14.34 | No clear effect |
Closest papers by meaning
- Occlusion-aware Non-Rigid Point Cloud Registration via Unsupervised Neural Deformation Correntropy (ICLR 2025 · similarity 0.88)
- PosDiffNet: Positional Neural Diffusion for Point Cloud Registration in a Large Field of View with Perturbations (AAAI 2024 · similarity 0.86)
- NeuroGauss4D-PCI: 4D Neural Fields and Gaussian Deformation Fields for Point Cloud Interpolation (NeurIPS 2024 · similarity 0.86)
- HybridReg: Robust 3D Point Cloud Registration with Hybrid Motions (AAAI 2025 · similarity 0.85)
- DV-Matcher: Deformation-based Non-rigid Point Cloud Matching Guided by Pre-trained Visual Features (CVPR 2025 · similarity 0.85)
- 4Deform: Neural Surface Deformation for Robust Shape Interpolation (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.
