Research map / Point clouds / Point cloud classification
SOE-Net: A Self-Attention and Orientation Encoding Network for Point Cloud Based Place Recognition
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 |
|---|---|---|---|
| Add self-attention to PointNetVLAD (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1%: 81.01 -> 86.71 | Helped |
| Add self-attention to PointNetVLAD (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1: 62.76 -> 73.03 | Helped |
| Add PointOE to PointNetVLAD (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1%: 81.01 -> 92.20 | Helped |
| Add PointOE to PointNetVLAD (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1: 62.76 -> 82.21 | Helped |
| Add PointOE and self-attention to PointNetVLAD (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1%: 81.01 -> 93.41 | Helped |
| Add PointOE and self-attention to PointNetVLAD (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1: 62.76 -> 84.20 | Helped |
| Replace PointNet with PointNet++ (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1%: 81.01 -> 89.10 | Helped |
| Replace PointNet with PointNet++ (ablation) | PointNetVLAD · Oxford RobotCar | average recall @1: 62.76 -> 76.23 | Helped |
| Replace HPHN quadruplet loss with lazy quadruplet loss (ablation) | SOE-Net · Oxford RobotCar | average recall @1%: 96.40 -> 93.41 | Hurt |
| Replace HPHN quadruplet loss with lazy quadruplet loss (ablation) | SOE-Net · Oxford RobotCar | average recall @1: 89.37 -> 84.20 | Hurt |
| Reduce global descriptor dimension from 256 to 128 (ablation) | SOE-Net · Oxford RobotCar | average recall @1%: 96.40 -> 95.30 | Hurt |
| Increase global descriptor dimension from 256 to 512 (ablation) | SOE-Net · Oxford RobotCar | average recall @1%: 96.40 -> 96.70 | Helped |
Closest papers by meaning
- PointNetVLAD: Deep Point Cloud Based Retrieval for Large-Scale Place Recognition (CVPR 2018 · similarity 0.91)
- Sonata: Self-Supervised Learning of Reliable Point Representations (CVPR 2025 · similarity 0.86)
- Mitigating Object Dependencies: Improving Point Cloud Self-Supervised Learning through Object Exchange (CVPR 2024 · similarity 0.85)
- LinNet: Linear Network for Efficient Point Cloud Representation Learning (NeurIPS 2024 · similarity 0.85)
- RI-MAE: Rotation-Invariant Masked AutoEncoders for Self-Supervised Point Cloud Representation Learning (AAAI 2025 · similarity 0.85)
- Improved MLP Point Cloud Processing with High-Dimensional Positional Encoding (AAAI 2024 · similarity 0.84)
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.