Research map / Point clouds / Shape and surface completion
GraphX-Convolution for Point Cloud Deformation in 2D-to-3D Conversion
Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sanghoon Lee
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 AdaIN global feature to projection-only model (ablation) | PCDNet · ShapeNet 5 categories | Chamfer distance: 0.452 -> 0.298 | Improved |
| add point projection feature to AdaIN-only model (ablation) | PCDNet · ShapeNet 5 categories | Chamfer distance: 0.421 -> 0.298 | Improved |
| add AdaIN global feature to projection-only model (ablation) | PCDNet · ShapeNet 5 categories | IoU: 0.644 -> 0.716 | Improved |
| add point projection feature to AdaIN-only model (ablation) | PCDNet · ShapeNet 5 categories | IoU: 0.651 -> 0.716 | Improved |
| replace FC deformation layers with GraphX (ablation) | PCDNet · ShapeNet 13 categories | Chamfer distance: 0.286 -> 0.262 | Improved |
| replace FC deformation layers with GraphX (ablation) | PCDNet · ShapeNet 13 categories | IoU: 0.743 -> 0.648 | Hurt |
| add residual connections to FC deformation network (ablation) | PCDNet · ShapeNet 13 categories | Chamfer distance: 0.286 -> 0.276 | Improved |
| add residual connections to GraphX (ablation) | PCDNet · ShapeNet 13 categories | Chamfer distance: 0.262 -> 0.259 | Improved |
| add residual connections to GraphX (ablation) | PCDNet · ShapeNet 13 categories | IoU: 0.648 -> 0.732 | Improved |
| add upsampling to ResGraphX (ablation) | PCDNet · ShapeNet 13 categories | Chamfer distance: 0.259 -> 0.252 | Improved |
| add upsampling to ResGraphX (ablation) | PCDNet · ShapeNet 13 categories | IoU: 0.732 -> 0.725 | Hurt |
| deform random point cloud with global and per-point features vs direct generation (PSG) | PCDNet vs PSG · ShapeNet 13 categories | Chamfer distance: 0.593 -> 0.286 | Improved |
Closest papers by meaning
- Efficient Dense Point Cloud Object Reconstruction using Deformation Vector Fields (ECCV 2018 · similarity 0.89)
- Convolution in the Cloud: Learning Deformable Kernels in 3D Graph Convolution Networks for Point Cloud Analysis (CVPR 2020 · similarity 0.89)
- Point Cloud Sampling via Graph Balancing and Gershgorin Disc Alignment (TPAMI 2022 · similarity 0.88)
- A Point Set Generation Network for 3D Object Reconstruction From a Single Image (CVPR 2017 · similarity 0.88)
- GEOMetrics: Exploiting Geometric Structure for Graph-Encoded Objects (ICML 2019 · similarity 0.88)
- SPU-PMD: Self-Supervised Point Cloud Upsampling via Progressive Mesh Deformation (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.
