Research map / Point clouds / Shape and surface completion
SPoVT: Semantic-Prototype Variational Transformer for Dense Point Cloud Semantic Completion
Sheng Yu Huang, Hao-Yu Hsu, Yu-Chiang Frank Wang
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 reconstruction pretrain to baseline model (ablation) | SPoVT · PCN Chair | mIoU: 43.07 -> 62.83 | Improved |
| add reconstruction pretrain to baseline model (ablation) | SPoVT · PCN Chair | Chamfer Distance: 5.23 -> 5.01 | Improved |
| add ratio predictor theta_P (ablation) | SPoVT · PCN Chair | mIoU: 62.83 -> 81.33 | Improved |
| add ratio predictor theta_P (ablation) | SPoVT · PCN Chair | Chamfer Distance: 5.01 -> 4.40 | Improved |
| add Semantic VAE (ablation) | SPoVT · PCN Chair | Chamfer Distance: 4.40 -> 2.99 | Improved |
| add Semantic VAE (ablation) | SPoVT · PCN Chair | mIoU: 81.33 -> 84.81 | Improved |
| add L_GCD loss (ablation) | SPoVT · PCN Chair | Chamfer Distance: 2.99 -> 2.36 | Improved |
| add L_GCD loss (ablation) | SPoVT · PCN Chair | mIoU: 84.81 -> 85.22 | Improved |
| use segmentation labels predicted by pre-trained DGCNN instead of ground truth (ablation) | SPoVT · PCN average | Chamfer Distance: 2.51 -> 2.85 | Hurt |
| use segmentation labels predicted by pre-trained DGCNN instead of ground truth (ablation) | SPoVT · PCN average | mIoU: 85.7 -> 81.2 | Hurt |
| train one unified model on Airplane and Car (ablation) | SPoVT · PCN Airplane | Chamfer Distance: 0.73 -> 0.84 | Hurt |
| train one unified model on Airplane and Car (ablation) | SPoVT · PCN Car | Chamfer Distance: 2.86 -> 3.39 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- PMP-Net++ — PMP-Net: Point Cloud Completion by Learning Multi-Step Point Moving Paths (CVPR 2021) · better on 2 of 2 shared results
- PoinTr — PoinTr: Diverse Point Cloud Completion With Geometry-Aware Transformers (ICCV 2021) · 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
- Point Cloud Semantic Scene Completion with Prototype-Guided Transformer (AAAI 2026 · similarity 0.91)
- Semantic Guided Part Relation-aware Network for Point Cloud Completion (AAAI 2026 · similarity 0.89)
- PoinTr: Diverse Point Cloud Completion With Geometry-Aware Transformers (ICCV 2021 · similarity 0.89)
- Position-Aware Guided Point Cloud Completion with CLIP Model (AAAI 2025 · similarity 0.87)
- SymmCompletion: High-Fidelity and High-Consistency Point Cloud Completion with Symmetry Guidance (AAAI 2025 · similarity 0.87)
- OneFormer3D: One Transformer for Unified Point Cloud Segmentation (CVPR 2024 · similarity 0.87)
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.
