Research map / 3D vision / 3D generation
TeethGenerator: A two-stage framework for paired pre- and post-orthodontic 3D dental data generation
Changsong Lei, Yaqian Liang, Shaofeng Wang, Jiajia Dai, Yong-Jin Liu
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 |
|---|---|---|---|
| use TeethGenerator instead of general point cloud generators (best baseline PVD) | Stage I generation · Post-orthodontic teeth generation (720 samples) | 1-NNA CD (%): 84.87 -> 69.50 | Improved |
| use TeethGenerator instead of DPM for uniqueness | Stage I generation · Post-orthodontic teeth generation (720 samples) | U_CD (%): 75.69 -> 96.25 | Improved |
| add voxel-based feature extraction (ablation) | TeethGenerator Stage I · Post-orthodontic teeth generation (720 samples) | U_CD (%): 21.11 -> 96.25 | Improved |
| use 4x4x4 instead of 3x3x3 voxels (ablation) | TeethGenerator Stage I · Post-orthodontic teeth generation (720 samples) | 1-NNA CD (%): 82.63 -> 69.50 | Improved |
| use 4x4x4 instead of 5x5x5 voxels (ablation) | TeethGenerator Stage I · Post-orthodontic teeth generation (720 samples) | 1-NNA CD (%): 77.75 -> 69.50 | Improved |
| use tooth-wise voxel partitioning instead of global 8x8x32 partitioning (ablation) | TeethGenerator Stage I · Post-orthodontic teeth generation (720 samples) | 1-NNA CD (%): 78.16 -> 69.50 | Improved |
| use 2x2x8 spatial structure instead of 1x1x32 (ablation) | TeethGenerator Stage I · Post-orthodontic teeth generation (720 samples) | 1-NNA CD (%): 74.48 -> 69.50 | Improved |
| add synthetic data to real training data for TANet, up to 10x real amount | TANet tooth alignment · TANet teeth arrangement (real 720 + synthetic) | ADD, PA-ADD, CSA | Improved |
| use style-only conditioning instead of applying ground-truth transformation parameters (ablation) | Stage II style transfer · Pre-orthodontic generation (qualitative) | visual quality | Improved |
| keep tooth-wise voxelization for style extractor instead of swapping with shape extractor (ablation) | Stage II · Pre-orthodontic generation (qualitative) | visual style fidelity | Improved |
Closest papers by meaning
- Collaborative Tooth Motion Diffusion Model in Digital Orthodontics (AAAI 2024 · similarity 0.89)
- Teeth-SEG: An Efficient Instance Segmentation Framework for Orthodontic Treatment based on Multi-Scale Aggregation and Anthropic Prior Knowledge (CVPR 2024 · similarity 0.85)
- Segmenting Maxillofacial Structures in CBCT Volumes (CVPR 2025 · similarity 0.84)
- NeBLa: Neural Beer-Lambert for 3D Reconstruction of Oral Structures from Panoramic Radiographs (AAAI 2024 · similarity 0.83)
- OralGPT-Omni: A Versatile Dental Multimodal Large Language Model (CVPR 2026 · similarity 0.83)
- OralGPT-Plus: Learning to Use Visual Tools via Reinforcement Learning for Panoramic X-ray Analysis (CVPR 2026 · similarity 0.82)
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.
