Research map / Point clouds / Registration and matching
Deep Confidence Guided Distance for 3D Partial Shape Registration
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 CGD instead of Chamfer distance to score candidate alignments (ablation) | CGD-net · ModelNet40 | RMSE(R): 11.53 -> 2.90; 8.63 | Helped |
| Use CGD instead of Chamfer distance to score candidate alignments (ablation) | CGD-net · FAUST | RMSE(R): 9.52 -> 3.81 | Helped |
| Add repulsion loss (ablation) | CGD-net · ModelNet40 | RMSE(R): 6.18 -> 2.90; 3.28 | Helped |
| Add repulsion loss (ablation) | CGD-net · FAUST | RMSE(R): 5.45 -> 3.81 | Helped |
| Add similarity loss (ablation) | CGD-net · ModelNet40 | RMSE(R): 5.39 -> 2.90 | Helped |
| Add similarity loss (ablation) | CGD-net · FAUST | RMSE(R): 5.13 -> 3.81 | Helped |
| Use the learned metric loss instead of direct transformation-parameter optimization (ablation) | CGD-net · ModelNet40 | RMSE(R): 6.79 -> 2.90 | Helped |
| Use the learned metric loss instead of direct transformation-parameter optimization (ablation) | CGD-net · FAUST | RMSE(R): 6.14 -> 3.81 | Helped |
| Use the hierarchical network with repulsion loss instead of DGCNN (ablation) | CGD-net · ModelNet40 | RMSE(R): 6.32 -> 2.90 | Helped |
| Use the hierarchical network with repulsion loss instead of DGCNN (ablation) | CGD-net · FAUST | RMSE(R): 5.88 -> 3.81 | Helped |
Closest papers by meaning
- PRNet: Self-Supervised Learning for Partial-to-Partial Registration (NeurIPS 2019 · similarity 0.90)
- Occlusion-aware Non-Rigid Point Cloud Registration via Unsupervised Neural Deformation Correntropy (ICLR 2025 · similarity 0.88)
- Non-Rigid Shape Registration via Deep Functional Maps Prior (NeurIPS 2023 · similarity 0.88)
- Partial-to-Partial Shape Matching with Geometric Consistency (CVPR 2024 · similarity 0.87)
- Scalable 3D Registration via Truncated Entry-wise Absolute Residuals (CVPR 2024 · similarity 0.87)
- Rectified Point Flow: Generic Point Cloud Pose Estimation (NeurIPS 2025 · similarity 0.87)
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.