Research map / 3D vision / Face and shape
View-GCN: View-Based Graph Convolutional Network for 3D Shape Analysis
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 local graph convolution (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 78.8 -> 79.8 | Helped |
| add non-local message passing (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 77.7 -> 79.8 | Helped |
| use selective view-sampling rather than FPS (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 78.2 -> 79.8 | Helped |
| learn the kNN affinity matrix (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 79.2 -> 79.8 | Helped |
| learn the kNN affinity matrix (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 79.1 -> 79.8 | Helped |
| use a three-level graph hierarchy rather than one level (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 78.5 -> 79.8 | Helped |
| use a three-level graph hierarchy rather than two levels (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 78.6 -> 79.8 | Helped |
| add view-selector loss (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 79.7 -> 79.8 | No clear effect |
| pool features after local graph convolution instead of after non-local message passing (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 78.3 -> 79.8 | Helped |
| increase sampling rate s to 0.6 (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 0.1% | No clear effect |
| increase sampling rate s to 0.7 (ablation) | View-GCN · ShapeNet Core55 | per class accuracy: 0.3% | Hurt |
| use the irregular view configuration with View-GCN instead of MVCNN-new (ablation) | 3D shape classification · irregular view configuration | per class accuracy: 4.2% | Helped |
Closest papers by meaning
- GVCNN: Group-View Convolutional Neural Networks for 3D Shape Recognition (CVPR 2018 · similarity 0.93)
- FeaStNet: Feature-Steered Graph Convolutions for 3D Shape Analysis (CVPR 2018 · similarity 0.89)
- Robust 3D Shape Classification via Non-Local Graph Attention Network (CVPR 2023 · similarity 0.88)
- Convolution in the Cloud: Learning Deformable Kernels in 3D Graph Convolution Networks for Point Cloud Analysis (CVPR 2020 · similarity 0.88)
- MLVCNN: Multi-Loop-View Convolutional Neural Network for 3D Shape Retrieval (AAAI 2019 · similarity 0.88)
- DGC-GNN: Leveraging Geometry and Color Cues for Visual Descriptor-Free 2D-3D Matching (CVPR 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.