Research map / 3D vision / 3D generation
Learning Unsupervised Hierarchical Part Decomposition of 3D Objects From a Single RGB Image
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 hierarchical superquadric part decomposition | Single-image reconstruction model · ShapeNet mean | IoU: 0.567 -> 0.580 | Helped |
| Use hierarchical superquadric part decomposition | Single-image reconstruction model · D-FAUST | IoU: 0.608 -> 0.699 | Helped |
| Use hierarchical superquadric part decomposition | Single-image reconstruction model · D-FAUST | Chamfer-L1: 0.189 -> 0.098 | Helped |
| Use unbiased importance sampling instead of uniform sampling (ablation) | Ours · ShapeNet chair test | IoU: 0.383 -> 0.491 | Helped |
| Use unbiased importance sampling instead of equal inside/outside sampling without reweighting (ablation) | Ours · ShapeNet chair test | IoU: 0.351 -> 0.491 | Helped |
| Use unbiased importance sampling instead of equal inside/outside sampling without reweighting (ablation) | Ours · ShapeNet chair test | Chamfer-L1: 0.041 -> 0.073 | Hurt |
| Increase importance-sampled points from 2k to 10k (ablation) | Ours · ShapeNet chair test | IoU: 0.370 -> 0.491 | Helped |
| Increase importance-sampled points from 5k to 10k (ablation) | Ours · ShapeNet chair test | IoU: 0.380 -> 0.491 | Helped |
| Add proximity loss to keep primitive centers near part centroids (ablation) | Ours · D-FAUST | IoU: 0.605 -> 0.699 | Helped |
| Add proximity loss to keep primitive centers near part centroids (ablation) | Ours · D-FAUST | Chamfer-L1: 0.171 -> 0.098 | Helped |
| Train decomposition and geometry in two optimization stages instead of jointly (ablation) | Ours · Not specified | Reconstruction quality | Hurt |
Closest papers by meaning
- Im2Struct: Recovering 3D Shape Structure From a Single RGB Image (CVPR 2018 · similarity 0.92)
- Neural Parts: Learning Expressive 3D Shape Abstractions With Invertible Neural Networks (CVPR 2021 · similarity 0.91)
- Learning Partonomic 3D Reconstruction from Image Collections (CVPR 2025 · similarity 0.91)
- Unsupervised Learning of 3D Structure from Images (NeurIPS 2016 · similarity 0.90)
- RIM-Net: Recursive Implicit Fields for Unsupervised Learning of Hierarchical Shape Structures (CVPR 2022 · similarity 0.90)
- PartCrafter: Structured 3D Mesh Generation via Compositional Latent Diffusion Transformers (NeurIPS 2025 · similarity 0.88)
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.