Research map / 3D vision / Face and shape
Persistent Homology Based Graph Convolution Network for Fine-Grained 3D Shape Segmentation
Chi-Chong Wong, Chi-Man Vong
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 persistent homology module in place of a general GCN module (ablation) | PHGCN · ShapeNet-Part | part-wise mIoU (%): 85.2 -> 89.2 | Improved |
| add the persistent homology module in place of a general GCN module (ablation) | PHGCN · PartNet level-3 | part-wise mIoU (%): 46.3 -> 49.8 | Improved |
| add the persistence diagram loss L_PD to the cross-entropy loss (ablation) | PHGCN · ShapeNet-Part | part-wise mIoU (%): 86.3 -> 89.2 | Improved |
| add the persistence diagram loss L_PD to the cross-entropy loss (ablation) | PHGCN · PartNet level-3 | part-wise mIoU (%): 47.0 -> 49.8 | Improved |
| add both PH module and L_PD to a general GCN trained with cross-entropy (ablation) | general GCN · ShapeNet-Part | part-wise mIoU (%): 83.4 -> 89.2 | Improved |
| add both PH module and L_PD to a general GCN trained with cross-entropy (ablation) | general GCN · PartNet level-3 | part-wise mIoU (%): 44.2 -> 49.8 | Improved |
| use PHGCN instead of ADConvnet | point cloud segmentation · ShapeNet-Part | part-wise mIoU (%): 87.0 -> 89.2 | Improved |
| use PHGCN instead of ADConvnet | point cloud segmentation · PartNet level-3 | part-wise mIoU (%): 47.0 -> 49.8 | Improved |
Closest papers by meaning
- TopoGAT: Plug-and-Play Topological Graph Attention for Fine-Grained 3D Segmentation (ECCV 2026 · similarity 0.89)
- Learning Fine-Grained Segmentation of 3D Shapes Without Part Labels (CVPR 2021 · similarity 0.88)
- A Topological Loss Function for Deep-Learning Based Image Segmentation Using Persistent Homology (TPAMI 2020 · similarity 0.88)
- Persistent Homology With Improved Locality Information for More Effective Delineation (TPAMI 2023 · similarity 0.87)
- Topograph: An Efficient Graph-Based Framework for Strictly Topology Preserving Image Segmentation (ICLR 2025 · similarity 0.87)
- SegGraph: Leveraging Graphs of SAM Segments for Few-Shot 3D Part Segmentation (NeurIPS 2025 · similarity 0.86)
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.
