Research map / 3D vision / 3D generation
Embedding 3D Geometric Features for Rigid Object Part Segmentation
Yafei Song, Xiaowu Chen, Jia Li, Qinping Zhao
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 |
|---|---|---|---|
| teach GeoNet with VolNet instead of random initialization (ablation) | Ours (2-stream) · PASCAL-Part Car | mean IoU (%): 57.01 -> 61.93 | Improved |
| teach GeoNet with VolNet instead of random initialization (ablation) | Ours (2-stream) · PASCAL-Part Aeroplane | mean IoU (%): 47.00 -> 50.50 | Improved |
| teach GeoNet with VolNet instead of random initialization (ablation) | Ours (2-stream) · PASCAL-Part Motorbike | mean IoU (%): 37.10 -> 38.66 | Improved |
| add pre-trained GeoNet stream to AppNet (DeepLab) | DeepLab · PASCAL-Part Car | mean IoU (%): 56.08 -> 61.93 | Improved |
| add pre-trained GeoNet stream to AppNet (DeepLab) | DeepLab · PASCAL-Part Aeroplane | mean IoU (%): 45.10 -> 50.50 | Improved |
| add pre-trained GeoNet stream to AppNet (DeepLab) | DeepLab · PASCAL-Part Motorbike | mean IoU (%): 37.07 -> 38.66 | Improved |
| add randomly initialized GeoNet stream to AppNet | DeepLab · PASCAL-Part Motorbike | mean IoU (%): 37.07 -> 37.10 | No clear effect |
| apply DCT compression to volume input of VolNet (ablation) | VolNet · ShapeNet Car | mean IoU (%): 62.66 -> 63.86 | Improved |
| apply DCT compression to volume input of VolNet (ablation) | VolNet · ShapeNet Aeroplane | mean IoU (%): 34.43 -> 36.37 | Improved |
| apply DCT compression to volume input of VolNet (ablation) | VolNet · ShapeNet Motorbike | mean IoU (%): 42.96 -> 44.99 | Improved |
| add GeoNet stream to AppNet on synthetic data | DeepLab · ShapeNet Aeroplane | mean IoU (%): 38.26 -> 44.99 | Improved |
| add GeoNet stream to AppNet on synthetic data | DeepLab · ShapeNet Motorbike | mean IoU (%): 52.46 -> 57.91 | Improved |
Closest papers by meaning
- Directionally Convolutional Networks for 3D Shape Segmentation (ICCV 2017 · similarity 0.88)
- Primal-Dual Mesh Convolutional Neural Networks (NeurIPS 2020 · similarity 0.88)
- DualConvMesh-Net: Joint Geodesic and Euclidean Convolutions on 3D Meshes (CVPR 2020 · similarity 0.87)
- Volumetric and Multi-View CNNs for Object Classification on 3D Data (CVPR 2016 · similarity 0.86)
- ImageNet3D: Towards General-Purpose Object-Level 3D Understanding (NeurIPS 2024 · similarity 0.85)
- SegGraph: Leveraging Graphs of SAM Segments for Few-Shot 3D Part Segmentation (NeurIPS 2025 · 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.
