Research map / Point clouds / Segmentation and detection
Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR-Based Perception
Xinge Zhu, Hui Zhou, Tai Wang, Fangzhou Hong, Wei Li, Yuexin Ma, Hongsheng Li, Ruigang Yang, Dahua Lin
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 |
|---|---|---|---|
| replace cubic partition with cylindrical partition (ablation) | 3D voxel baseline · SemanticKITTI (val) | mIoU: 58.1 -> 61.0 | Improved |
| add asymmetrical residual block networks (ablation) | Cylinder baseline · SemanticKITTI (val) | mIoU: 61.0 -> 63.8 | Improved |
| add dimension-decomposition context modeling (DDCM) (ablation) | Cylinder + Asym-CNN · SemanticKITTI (val) | mIoU: 63.8 -> 65.2 | Improved |
| add point-wise refinement module (ablation) | Cylinder + Asym-CNN + DDCM · SemanticKITTI (val) | mIoU: 65.2 -> 65.9 | Improved |
| use 1D-asymmetrical block instead of regular residual block (ablation) | Cylinder3D · SemanticKITTI (val) | mIoU: 60.8 -> 61.9 | Improved |
| use full asymmetrical block (with height) instead of regular block (ablation) | Cylinder3D · SemanticKITTI (val) | mIoU: 60.8 -> 63.9 | Improved |
| use 480x360x32 instead of 320x240x24 representation (ablation) | Cylinder3D · SemanticKITTI (val) | mIoU: 63.4 -> 65.9 | Improved |
| use 480x360x32 instead of 600x480x40 representation (ablation) | Cylinder3D · SemanticKITTI (val) | mIoU: 64.1 -> 65.9 | Improved |
| use cylindrical 480x360x32 instead of cubic 1000x1000x45 (ablation) | Cylinder3D · SemanticKITTI (val) | mIoU: 58.1 -> 65.9 | Improved |
| add cylindrical partition to SECOND (ablation) | SECOND · nuScenes detection | NDS: 46.8 -> 49.6 | Improved |
| add asymmetrical 3D convolution to SECOND (ablation) | SECOND · nuScenes detection | NDS: 46.8 -> 48.3 | Improved |
| add CyAs to SECOND | SECOND · nuScenes detection | mAP: 31.6 -> 36.4 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PolarNet — PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation (CVPR 2020) · better on 3 of 3 shared results
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
- Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation (CVPR 2021 · similarity 0.98)
- Less Is More: Reducing Task and Model Complexity for 3D Point Cloud Semantic Segmentation (CVPR 2023 · similarity 0.91)
- SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation (ECCV 2020 · similarity 0.91)
- Large-Scale Point Cloud Semantic Segmentation With Superpoint Graphs (CVPR 2018 · similarity 0.91)
- LiDAR-Based Panoptic Segmentation via Dynamic Shifting Network (CVPR 2021 · similarity 0.91)
- Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning Contextual Shape Priors from Scene Completion (AAAI 2021 · similarity 0.91)
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.
