Research map / Point clouds / Registration and matching
Deep Hough Voting for Robust Global Registration
Junha Lee, Seungwook Kim, Minsu Cho, Jaesik Park
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 learned Hough space refinement (ablation) | Hough voting registration · 3DMatch (5cm voxel) | recall (%): 86.7 -> 91.4 | Improved |
| add learned Hough space refinement (ablation) | Hough voting registration · 3DMatch (2.5cm voxel) | recall (%): 90.5 -> 93.7 | Improved |
| add Gaussian smoothing after voting (ablation) | Hough voting registration · 3DMatch (5cm voxel) | recall (%): 86.7 -> 89.2 | Improved |
| add Gaussian smoothing after voting (ablation) | Hough voting registration · 3DMatch (2.5cm voxel) | recall (%): 90.5 -> 91.7 | Improved |
| add Gaussian smoothing on top of learned refinement (ablation) | Hough voting with refinement · 3DMatch (5cm voxel) | recall (%): 91.4 -> 91.5 | No clear effect |
| add Gaussian smoothing on top of learned refinement (ablation) | Hough voting with refinement · 3DMatch (2.5cm voxel) | recall (%): 93.7 -> 93.1 | Hurt |
| use smaller voxel size 2.5cm instead of 5cm (ablation) | Hough voting with refinement · 3DMatch | recall (%): 91.4 -> 93.7 | Improved |
| use learned refinement instead of Gaussian smoothing (ablation) | Hough voting registration · 3DMatch (5cm voxel) | recall (%): 89.2 -> 91.4 | Improved |
| add learned refinement module (ablation) | Hough voting registration · 3DMatch | peak memory (GB): 2.949 -> 4.344 | Hurt |
| replace RANSAC-2M with Hough voting | registration · 3DMatch | recall (%): 66.1 -> 91.4 | Improved |
| replace DGR with Hough voting | registration · 3DMatch | recall (%): 85.2 -> 91.4 | Improved |
| replace PointDSC with Hough voting using Predator features | registration · 3DLoMatch | recall (%): 61.5 -> 64.6 | Improved |
Closest papers by meaning
- GeoTransformer: Fast and Robust Point Cloud Registration With Geometric Transformer (TPAMI 2023 · similarity 0.89)
- Deterministic Point Cloud Registration via Novel Transformation Decomposition (CVPR 2022 · similarity 0.89)
- Efficient and Robust Point Cloud Registration via Heuristics-Guided Parameter Search (TPAMI 2024 · similarity 0.88)
- Geometric Transformer for Fast and Robust Point Cloud Registration (CVPR 2022 · similarity 0.88)
- Register Anything Model for Generalizable and Robust Point Cloud Registration (NeurIPS 2026 · similarity 0.88)
- Distinctiveness Oriented Positional Equilibrium for Point Cloud Registration (ICCV 2021 · similarity 0.88)
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.
