Research map / 3D vision / Face and shape
Neural Star Domain as Primitive Representation
Yuki Kawana, Yusuke Mukuta, Tatsuya Harada
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 neural star-domain primitives instead of the strongest reported baseline | single-view 3D reconstruction · ShapeNet, 13-category mean | F-score: 53.51 -> 54.29 | Improved |
| use neural star-domain primitives instead of the strongest reported baseline | single-view 3D reconstruction · ShapeNet, 13-category mean | CD1: 0.189 -> 0.173 | Improved |
| use neural star-domain primitives instead of the strongest reported baseline | single-view 3D reconstruction · ShapeNet, 13-category mean | IoU: 0.593 -> 0.589 | Hurt |
| use explicit neural star-domain mesh sampling instead of OccNet* | mesh generation · ShapeNet, 13-category mean | mesh sampling time (seconds): 0.529 -> 0.014 | Improved |
| add surface point extraction to NSD10 (ablation) | NSDN · ShapeNet, F-score ablation | F-score: 45.84 -> 50.52 | Improved |
| add occupancy loss alongside surface point loss (ablation) | NSDN · ShapeNet reconstruction | reconstruction accuracy | Improved |
| use only occupancy loss instead of the combined reconstruction losses (ablation) | NSDN · ShapeNet reconstruction | reconstruction success | Hurt |
| add overlap regularization (ablation) | NSDN primitives · ShapeNet airplane and chair | part IoU | Improved |
| add overlap regularization (ablation) | NSDN primitives · ShapeNet airplane and chair | F-score | No clear effect |
Closest papers by meaning
- Neural Parts: Learning Expressive 3D Shape Abstractions With Invertible Neural Networks (CVPR 2021 · similarity 0.89)
- 3D-PRNN: Generating Shape Primitives With Recurrent Neural Networks (ICCV 2017 · similarity 0.87)
- NeuralPlane: Structured 3D Reconstruction in Planar Primitives with Neural Fields (ICLR 2025 · similarity 0.87)
- 3DIAS: 3D Shape Reconstruction With Implicit Algebraic Surfaces (ICCV 2021 · similarity 0.86)
- Learning Unsupervised Hierarchical Part Decomposition of 3D Objects From a Single RGB Image (CVPR 2020 · similarity 0.86)
- Learning Partonomic 3D Reconstruction from Image Collections (CVPR 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.
