Research map / Point clouds / Point cloud classification
Resolution-Free Point Cloud Sampling Network with Data Distillation
Tianxin Huang, Jiangning Zhang, Jun Chen, Yuang Liu, Yong Liu
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 resolution-aware pooling (ablation) | PCDNet · Ablation (AE) | CD: 3.70 -> 2.99 | Improved |
| add data distillation (ablation) | PCDNet · Ablation (AE) | CD: 2.99 -> 2.89 | Improved |
| add dynamic resolution selection (ablation) | PCDNet · Ablation (AE) | CD: 2.89 -> 2.74 | Improved |
| add move constraint (ablation) | PCDNet · Ablation (AE) | CD: 2.74 -> 2.67 | Improved |
| add all modules vs base on Hausdorff distance (ablation) | PCDNet · Ablation (AE) | HD: 3.31 -> 1.97 | Improved |
| compare PCDNet to S-Net and SamNet on cost | PCDNet · Efficiency (AE) | inference time (ms): 17 -> 7 | Improved |
| use random sampling instead of FPS for initial seeds (ablation) | PCDNet · Sensitivity (AE) | reconstruction error | Hurt |
| use Voxel or Sphere filtering instead of FPS for initial seeds (ablation) | PCDNet · Sensitivity (AE) | reconstruction error | No clear effect |
| add data distillation, effect larger at low resolution (ablation) | PCDNet · Recognition and reconstruction | task performance | Improved |
| use too small or too large resolution updating interval (ablation) | PCDNet · Ablation (AE) | CD and HD | Hurt |
| use too large or too small distillation margin m or temperature T (ablation) | PCDNet · Distillation hyper-parameters | task performance | Hurt |
| sample at 1024 instead of 512 points | PCDNet · ModelNet40 (PointNet) | accuracy (%): 86.32% at 512 points and 86.28% at 1024 points | No clear effect |
Closest papers by meaning
- GCD-Sampling: A General Cross-scale Decoupled Sampling for Point Cloud (AAAI 2025 · similarity 0.89)
- LTA-PCS: Learnable Task-Agnostic Point Cloud Sampling (CVPR 2024 · similarity 0.88)
- SPU-IMR: Self-supervised Arbitrary-scale Point Cloud Upsampling via Iterative Mask-recovery Network (AAAI 2025 · similarity 0.86)
- Point Cloud Dataset Distillation (ICML 2025 · similarity 0.86)
- Point Cloud Upsampling Using Conditional Diffusion Module with Adaptive Noise Suppression (CVPR 2025 · similarity 0.85)
- MICAS: Multi-grained In-Context Adaptive Sampling for 3D Point Cloud Processing (CVPR 2025 · 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.
