Research map / Point clouds / Segmentation and detection
PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation
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 |
|---|---|---|---|
| Use polar BEV instead of Cartesian BEV | Unet · SemanticKITTI test | mIoU: 50.7% -> 54.3% | Helped |
| Use polar BEV instead of Cartesian BEV | Unet · A2D2 test | mIoU: 20.3% -> 23.9% | Helped |
| Use polar BEV instead of Cartesian BEV | Unet · Paris-Lille-3D test | mIoU: 40.3% -> 43.7% | Helped |
| Add ring convolution (ablation) | Polar BEV Unet on SemanticKITTI val · SemanticKITTI val | mIoU: 46.9% -> 47.4% | Helped |
| Add 9 point features (ablation) | Polar BEV Unet on SemanticKITTI val · SemanticKITTI val | mIoU: 47.4% -> 48.5% | Helped |
| Add flip augmentation (ablation) | Polar BEV Unet on SemanticKITTI val · SemanticKITTI val | mIoU: 48.5% -> 50.6% | Helped |
| Fix the BEV volume space (ablation) | Polar BEV Unet on SemanticKITTI val · SemanticKITTI val | mIoU: 50.6% -> 53.4% | Helped |
| Tune grid size to [480,360,32] (ablation) | Polar BEV Unet on SemanticKITTI val · SemanticKITTI val | mIoU: 53.4% -> 54.9% | Helped |
| Use polar BEV instead of Cartesian BEV (ablation) | SqueezeSeg · SemanticKITTI val | mIoU: 42.6% -> 42.2% | No clear effect |
| Use polar BEV instead of Cartesian BEV (ablation) | Resnet-FCN · SemanticKITTI val | mIoU: 49.2% -> 52.5% | Helped |
| Use polar BEV instead of Cartesian BEV (ablation) | DRN-DL · SemanticKITTI val | mIoU: 46.7% -> 51.2% | Helped |
| Use polar BEV instead of Cartesian BEV (ablation) | Resnet-DL · SemanticKITTI val | mIoU: 50.4% -> 53.6% | Helped |
Closest papers by meaning
- Panoptic-PolarNet: Proposal-Free LiDAR Point Cloud Panoptic Segmentation (CVPR 2021 · similarity 0.93)
- LiDAR-Based Panoptic Segmentation via Dynamic Shifting Network (CVPR 2021 · similarity 0.91)
- PC-BEV: An Efficient Polar-Cartesian BEV Fusion Framework for LiDAR Semantic Segmentation (AAAI 2025 · similarity 0.90)
- HiLoTs: High-Low Temporal Sensitive Representation Learning for Semi-Supervised LiDAR Segmentation in Autonomous Driving (CVPR 2025 · similarity 0.88)
- Multi-Space Alignments Towards Universal LiDAR Segmentation (CVPR 2024 · similarity 0.88)
- No Thing, Nothing: Highlighting Safety-Critical Classes for Robust LiDAR Semantic Segmentation in Adverse Weather (CVPR 2025 · 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.