Research map / Point clouds / Shape and surface completion
Variational Relational Point Completion Network for Robust 3D Classification
Liang Pan, Xinyi Chen, Zhongang Cai, Junzhe Zhang, Haiyu Zhao, Shuai Yi, Ziwei 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 the proposed PSA kernel, dual-path architecture and PSK modules (ablation) | ECG baseline · MVP completion (2048 points) | Chamfer distance x10^4: 6.64 -> 5.96 | Improved |
| add the proposed modules (F-score) (ablation) | ECG baseline · MVP completion (2048 points) | F-score@1%: 0.476 -> 0.499 | Improved |
| add relational structure learning (VRCNet vs ECG) at 16,384 points | ECG · MVP completion (16,384 points) | Chamfer distance x10^4: 3.58 -> 3.06 | Improved |
| increase output resolution from 2048 to 16384 points | VRCNet · MVP completion | Chamfer distance x10^4: 5.96 -> 3.12 | Improved |
| complete partial clouds with VRCNet before classification, 50% missing | PointNet/DGCNN · MVP-40 classification | accuracy | Improved |
| complete partial clouds with PCN before DGCNN classification (25% missing) | DGCNN · MVP-40 classification | Avg accuracy: 69.0 -> 59.9 | Hurt |
| increase missing ratio from 25% to 50% | VRCNet · MVP-40 completion | Chamfer distance x10^3: 1.89 -> 2.76 | Hurt |
| replace VRCNet with PCN on real car scans in user study | VRCNet · KITTI car user study | preference score: 2.07 -> 2.87 | Hurt |
| use VRCNet on real chair scans vs PCN | PCN · ScanNet chair user study | preference score: 1.60 -> 2.67 | Improved |
| partial instead of complete input to classifier | PointNet and DGCNN · MVP classification | accuracy | Hurt |
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.
- GRNet — GRNet: Gridding Residual Network for Dense Point Cloud Completion (ECCV 2020) · better on 2 of 2 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
- Variational Relational Point Completion Network (CVPR 2021 · similarity 0.99)
- DC-PCN: Point Cloud Completion Network with Dual-Codebook Guided Quantization (AAAI 2025 · similarity 0.90)
- PointCFormer: A Relation-Based Progressive Feature Extraction Network for Point Cloud Completion (AAAI 2025 · similarity 0.90)
- ASHF-Net: Adaptive Sampling and Hierarchical Folding Network for Robust Point Cloud Completion (AAAI 2021 · similarity 0.90)
- CDPNet: Cross-Modal Dual Phases Network for Point Cloud Completion (AAAI 2024 · similarity 0.88)
- Digging into Intrinsic Contextual Information for High-fidelity 3D Point Cloud Completion (AAAI 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.
