Research map / Point clouds / Segmentation and detection
Unsupervised Feature Learning for 3D Scene Reconstruction with Occupancy Maps
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 |
|---|---|---|---|
| orientation clustering reduces cluster count from initial clusters z to planar clusters w | PS-HM · Room | number of clusters: 8120 -> 42 | Helped |
| orientation clustering reduces cluster count from initial clusters z to planar clusters w | PS-HM · Corridor | number of clusters: 7637 -> 83 | Helped |
| orientation clustering reduces cluster count from initial clusters z to planar clusters w | PS-HM · Outdoor | number of clusters: 10560 -> 23 | Mixed |
| clustering reduces points to initial clusters | ASK-means · Room | number of elements: 582463 -> 8120 | Helped |
| PS-HM versus LARD-HM and Octo-Map in structured environments | Hilbert Maps · Room and Corridor | classification accuracy | Helped |
| PS-HM in predominantly non-planar outdoor data | PS-HM · Outdoor | classification accuracy | No clear effect |
| add anisotropic nearest-neighbor search with length-scale weighting instead of Euclidean kd-tree (ablation) | PS-HM · sparse clusters | neighbor association quality | Helped |
| add 3D orientation alignment step after orientation clustering (ablation) | orientation clustering · 3D extruded dataset | wall rotation correctness | Helped |
| add unoccupied points sampled between sensor and occupied points (ablation) | orientation clustering · Room | reconstruction detail | Helped |
| orientation clustering step added to spatial clustering | preprocessing · all datasets | processing time: 20% | Hurt |
Closest papers by meaning
- Learning Occupancy for Monocular 3D Object Detection (CVPR 2024 · similarity 0.88)
- LowRankOcc: Tensor Decomposition and Low-Rank Recovery for Vision-based 3D Semantic Occupancy Prediction (CVPR 2024 · similarity 0.86)
- AdaOcc: Adaptive 3D Occupancy Prediction for Embodied Tasks (NeurIPS 2026 · similarity 0.86)
- Unsupervised Occupancy Learning from Sparse Point Cloud (CVPR 2024 · similarity 0.86)
- OctOcc: High-Resolution 3D Occupancy Prediction with Octree (AAAI 2024 · similarity 0.86)
- ODG: Occupancy Prediction Using Dual Gaussians (NeurIPS 2025 · similarity 0.86)
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.