Research map / 3D vision / Face and shape
TSGCNet: Discriminative Geometric Feature Learning With Two-Stream Graph Convolutional Network for 3D Dental Model Segmentation
Lingming Zhang, Yue Zhao, Deyu Meng, Zhiming Cui, Chenqiang Gao, Xinbo Gao, Chunfeng Lian, Dinggang Shen
Check 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 |
|---|---|---|---|
| use separate coordinate and normal-vector streams instead of one stream on concatenated inputs (ablation) | TSGCNet · 3D dental model test set | OA: 87.25 -> 95.25 | Improved |
| use separate coordinate and normal-vector streams instead of one stream on concatenated inputs (ablation) | TSGCNet · 3D dental model test set | mIoU: 73.44 -> 88.99 | Improved |
| add the N-stream to the coordinate-only C-stream (ablation) | TSGCNet-C · 3D dental model test set | OA: 83.23 -> 95.25 | Improved |
| add the C-stream to the normal-only N-stream (ablation) | TSGCNet-N · 3D dental model test set | mIoU: 20.77 -> 88.99 | Improved |
| use attention in the C-stream and max-pooling in the N-stream rather than max-pooling in both (ablation) | TSGCNet · 3D dental model test set | mIoU: 86.24 -> 88.99 | Improved |
| use max-pooling in the N-stream rather than attention, with attention in the C-stream (ablation) | TSGCNet · 3D dental model test set | mIoU: 87.35 -> 88.99 | Improved |
| use attention in the C-stream rather than max-pooling, with attention in the N-stream (ablation) | TSGCNet · 3D dental model test set | mIoU: 85.67 -> 87.35 | Improved |
| fuse the streams after separate feature extraction rather than at each layer (ablation) | TSGCNet · 3D dental model test set | mIoU: 85.49 -> 88.99 | Improved |
| replace MeshSegNet with TSGCNet | 3D dental model segmentation · 3D dental model test set | OA: 93.11 -> 95.25 | Improved |
| replace MeshSegNet with TSGCNet | 3D dental model segmentation · 3D dental model test set | mIoU: 84.47 -> 88.99 | Improved |
Later papers that compared with this method
2 later accepted papers report TSGCNet as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Photo-Guided Tooth Segmentation on 3D Oral Scan Model (CVPR 2026 · better on 1 of 1 shared result)
- 3DTeethSAM: Taming SAM2 for 3D Teeth Segmentation (AAAI 2026 · better on 1 of 1 shared result)
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
- Segmenting Maxillofacial Structures in CBCT Volumes (CVPR 2025 · similarity 0.88)
- 3D Dental Model Segmentation with Geometrical Boundary Preserving (CVPR 2025 · similarity 0.86)
- Teeth-SEG: An Efficient Instance Segmentation Framework for Orthodontic Treatment based on Multi-Scale Aggregation and Anthropic Prior Knowledge (CVPR 2024 · similarity 0.85)
- E2ENet: Dynamic Sparse Feature Fusion for Accurate and Efficient 3D Medical Image Segmentation (NeurIPS 2024 · similarity 0.83)
- Upping the Game: How 2D U-Net Skip Connections Flip 3D Segmentation (NeurIPS 2024 · similarity 0.83)
- SegGraph: Leveraging Graphs of SAM Segments for Few-Shot 3D Part Segmentation (NeurIPS 2025 · similarity 0.83)
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.
