Research map / Point clouds / Registration and matching
DeepMapping: Unsupervised Map Estimation From Multiple Point Clouds
Check 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 |
|---|---|---|---|
| Optimize network parameters with DeepMapping rather than optimizing sensor poses directly | point-cloud registration · 2D simulated point cloud, 128 scans | success rate: 31.5% -> 84.2% | Helped |
| Use DeepMapping instead of incremental ICP | point-cloud registration · 2D simulated point cloud, 128 scans | success rate: 36.0% -> 84.2%; 53.3% -> 84.2% | Helped |
| Use DeepMapping instead of multiway registration | point-cloud registration · Active Vision Dataset, 16 point clouds | success rate: 58.1% -> 80.0% | Helped |
| Add CNN feature extraction to the L-Net instead of PointNet (ablation) | DeepMapping · Active Vision Dataset | ATE: 207.84mm -> 134.07mm | Helped |
| Replace MLP with ResMLP in the M-Net (ablation) | DeepMapping · Active Vision Dataset | ATE | Hurt |
| Increase M-Net depth from 4 or 5 layers to 6 layers (ablation) | DeepMapping · Active Vision Dataset | ATE | Helped |
| Increase M-Net depth beyond 6 layers (ablation) | DeepMapping · Active Vision Dataset | ATE | Mixed |
| Increase M-Net width to 128 (ablation) | DeepMapping · Active Vision Dataset | ATE | Helped |
| Add Chamfer distance to the BCE loss (ablation) | DeepMapping · Active Vision Dataset | registration accuracy | Helped |
| Use network-based optimization rather than direct gradient descent on the original variable | non-convex 1D optimization example · 1D non-convex objective | ability to reach the optimum | Helped |
Closest papers by meaning
- DeepPointMap: Advancing LiDAR SLAM with Unified Neural Descriptors (AAAI 2024 · similarity 0.93)
- Unlocking Generalization Power in LiDAR Point Cloud Registration (CVPR 2025 · similarity 0.88)
- GeoNLF: Geometry guided Pose-Free Neural LiDAR Fields (NeurIPS 2024 · similarity 0.87)
- LiDAR-Net: A Real-scanned 3D Point Cloud Dataset for Indoor Scenes (CVPR 2024 · similarity 0.87)
- A Consistency-Aware Spot-Guided Transformer for Versatile and Hierarchical Point Cloud Registration (NeurIPS 2024 · similarity 0.87)
- FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimators (ICLR 2024 · 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.