Research map / 3D vision / Face and shape
Learning Mesh Representations via Binary Space Partitioning Tree Networks
Zhiqin Chen, Andrea Tagliasacchi, Hao Zhang
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 BSP-Net instead of BAE | shape decomposition · ShapeNet auto-encoding | Chamfer Distance: 1.592 -> 0.447 | Improved |
| use BSP-Net instead of BAE | shape decomposition · ShapeNet auto-encoding | Light Field Distance: 4587.34 -> 2019.26 | Improved |
| add overlap loss (ablation) | BSP-Net · ShapeNet auto-encoding | Chamfer Distance: 0.447 -> 0.448 | No clear effect |
| add overlap loss (ablation) | BSP-Net · ShapeNet Part segmentation | mean IoU: 79.3 -> 79.8 | Improved |
| use BSP-Net instead of BAE (one-shot) | segmentation · ShapeNet Part | mean IoU: 59.8 -> 79.3 | Improved |
| add soft threshold loss L*_T (ablation) | BSP-Net · shape auto-encoding | Chamfer Distance (before): 0.743 -> 0.735 | Improved |
| add L*_T and overlap loss (ablation) | BSP-Net · shape auto-encoding | Chamfer Distance (before): 0.743 -> 0.724 | Improved |
| add L*_T and overlap loss (ablation) | BSP-Net · single view reconstruction | Chamfer Distance (before): 1.432 -> 1.453 | No clear effect |
| vary threshold lambda (ablation) | BSP-Net · shape auto-encoding | Chamfer Distance: 0.743 -> 0.758 | No clear effect |
| use BSP-Net instead of Atlas25 for sharp edges | single view reconstruction · ShapeNet SVR | Edge Chamfer Distance: 2.069 -> 0.743 | Improved |
| use BSP-Net instead of Atlas25 | single view reconstruction · ShapeNet SVR | Chamfer Distance: 1.170 -> 1.432 | Hurt |
| use BSP-Net instead of IM-NET256 for mesh size | single view reconstruction · ShapeNet SVR | vertex count: 82965 -> 1073 | Improved |
Closest papers by meaning
- BSP-Net: Generating Compact Meshes via Binary Space Partitioning (CVPR 2020 · similarity 0.98)
- CvxNet: Learnable Convex Decomposition (CVPR 2020 · similarity 0.89)
- Neural Parts: Learning Expressive 3D Shape Abstractions With Invertible Neural Networks (CVPR 2021 · similarity 0.89)
- SkeletonNet: A Topology-Preserving Solution for Learning Mesh Reconstruction of Object Surfaces From RGB Images (TPAMI 2021 · similarity 0.88)
- NeuralSlice: Neural 3D Triangle Mesh Reconstruction via Slicing 4D Tetrahedral Meshes (ICML 2023 · similarity 0.87)
- MeshGPT: Generating Triangle Meshes with Decoder-Only Transformers (CVPR 2024 · similarity 0.87)
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.
