Research map / Point clouds / Point cloud classification
Point Cloud Compression with Range Image-Based Entropy Model for Autonomous Driving
Sukai Wang, Ming Liu
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 stage 0 (non-ground split) to the entropy model (ablation) | RICNet · q1=0.2 | BPP: 2.04 -> 2.04 | Improved |
| add stage 0 (non-ground split) to the entropy model (ablation) | RICNet · q1=0.6 | BPP: 2.18 -> 2.13 | Improved |
| add stages 0 and 1 (entropy model) over BZip2 compressor (ablation) | RICNet · q1=0.3 | BPP: 2.27 -> 1.92 | Improved |
| use q1/q2 = 3 in the first quantizer (ablation) | RICNet · BPP vs q1 | BPP: 1.92 | Improved |
| add 2D conv fusion to Minkowski 3D extractor (ablation) | RICNet · KITTI city | BPP: 2.38 -> 2.26 | Improved |
| add 2D conv fusion to Minkowski 3D extractor (ablation) | RICNet · KITTI city | F1 score: 0.467 -> 0.466 | No clear effect |
| replace 2D conv with attentive conv (SAC) (ablation) | RICNet · KITTI city | BPP: 2.26 -> 2.25 | Improved |
| replace 2D conv with attentive conv (SAC) (ablation) | RICNet · KITTI city | F1 score: 0.466 -> 0.508 | Improved |
| replace 2D conv with attentive conv (SAC) (ablation) | RICNet · KITTI city | Chamfer distance: 0.027 -> 0.0255 | Improved |
| use Minkowski UNet instead of PointNet++ (ablation) | RICNet · KITTI city | BPP: 2.46 -> 2.38 | Improved |
| use Minkowski UNet instead of SqueezeSeg (ablation) | RICNet · KITTI city | F1 score: 0.285 -> 0.467 | Improved |
Closest papers by meaning
- RIDDLE: Lidar Data Compression With Range Image Deep Delta Encoding (CVPR 2022 · similarity 0.91)
- OctSqueeze: Octree-Structured Entropy Model for LiDAR Compression (CVPR 2020 · similarity 0.90)
- SCP: Spherical-Coordinate-Based Learned Point Cloud Compression (AAAI 2024 · similarity 0.86)
- msLPCC: A Multimodal-Driven Scalable Framework for Deep LiDAR Point Cloud Compression (AAAI 2024 · similarity 0.85)
- RENO: Real-Time Neural Compression for 3D LiDAR Point Clouds (CVPR 2025 · similarity 0.85)
- BEV-MAE: Bird’s Eye View Masked Autoencoders for Point Cloud Pre-training in Autonomous Driving Scenarios (AAAI 2024 · similarity 0.85)
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.
