Research map / Point clouds / Point cloud classification
BEVPlace: Learning LiDAR-based Place Recognition using Bird's Eye View Images
Lun Luo, Shuhang Zheng, Yixuan Li, Yongzhi Fan, Beinan Yu, Si-Yuan Cao, Junwei Li, Hui-Liang Shen
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 group convolution rotation invariance to BEV NetVLAD (ablation) | ResNet34+NetVLAD · KITTI | mean Recall@1: 92.8 -> 99.3 | Improved |
| add group convolution rotation invariance to BEV NetVLAD (ablation) | ResNet34+NetVLAD · KITTI rotated test | mean Recall@1: 84.6 -> 98.0 | Improved |
| add group convolution rotation invariance to BEV NetVLAD (ablation) | ResNet34+NetVLAD · Benchmark dataset | mean AR@1: 94.2 -> 95.3 | Improved |
| add group convolution rotation invariance to BEV NetVLAD (ablation) | ResNet34+NetVLAD · Benchmark dataset | mean AR@1%: 99.4 -> 99.2 | No clear effect |
| use BEV image representation with plain NetVLAD instead of unordered points (PointNetVLAD) (ablation) | PointNetVLAD · KITTI | mean Recall@1: 77.2 -> 92.8 | Improved |
| use BEV with plain NetVLAD instead of OverlapTransformer range images (ablation) | OverlapTransformer · KITTI | mean Recall@1: 91.1 -> 92.8 | Improved |
| use ResNet34 instead of ResNet18 backbone (ablation) | BEV NetVLAD · KITTI | mean Recall@1: 92.0 -> 92.8 | Improved |
| replace PointNet with Vector Neuron backbone for rotation invariance (ablation) | PointNetVLAD · KITTI rotated test | mean Recall@1: 62.1 -> 80.9 | Improved |
| replace PointNet with Vector Neuron backbone for rotation invariance (ablation) | PointNetVLAD · KITTI | mean Recall@1: 77.2 -> 82.9 | Improved |
| rotate test point clouds around z-axis | PointNetVLAD · KITTI | mean Recall@1: 77.2 -> 62.1 | Hurt |
| rotate test point clouds around z-axis | BEVPlace · KITTI | mean Recall@1: 99.3 -> 98.0 | Hurt |
| rotate test point clouds around z-axis | ResNet34+NetVLAD · KITTI | mean Recall@1: 92.8 -> 84.6 | Hurt |
Later papers that compared with this method
2 later accepted papers report BEVPlace++ as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- C-LaV: Conditional Latent Velocity Field Denoising for Weather-Robust LiDAR Place Recognition (CVPR 2026 · better on 2 of 2 shared results)
- ForestLPR: LiDAR Place Recognition in Forests Attentioning Multiple BEV Density Images (CVPR 2025 · better on 0 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- ForestLPR: LiDAR Place Recognition in Forests Attentioning Multiple BEV Density Images (CVPR 2025 · similarity 0.89)
- Map It Anywhere: Empowering BEV Map Prediction using Large-scale Public Datasets (NeurIPS 2024 · similarity 0.88)
- L2RSI: Cross-view LiDAR-based Place Recognition for Large-scale Urban Scenes via Remote Sensing Imagery (NeurIPS 2025 · similarity 0.88)
- HOTFormerLoc: Hierarchical Octree Transformer for Versatile Lidar Place Recognition Across Ground and Aerial Views (CVPR 2025 · similarity 0.87)
- Multi-Modal Aerial-Ground Cross-View Place Recognition with Neural ODEs (CVPR 2025 · similarity 0.87)
- GeoBEV: Learning Geometric BEV Representation for Multi-view 3D Object Detection (AAAI 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.
