Research map / 3D vision / Face and shape
Residual Primitive Fitting of 3D Shapes with SuperFrusta
Aditya Ganeshan, Matheus Gadelha, Thibault Groueix, Zhiqin Chen, Siddhartha Chaudhuri, Vladimir Kim, Wang Yifan, Daniel Ritchie
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 ResFit + SuperFrustum instead of MPS | primitive fitting · 3DGen-Prim | IoU: 82.67 -> 88.74 | Improved |
| use ResFit + SuperFrustum instead of MPS | primitive fitting · Toys4K | IoU: 80.60 -> 89.92 | Improved |
| use ResFit + SuperFrustum instead of MPS | primitive fitting · 3DGen-Prim | #Prims: 42.96 -> 23.98 | Improved |
| use ResFit + SuperFrustum instead of MPS | primitive fitting · 3DGen-Prim | Overlap: 0.684 -> 0.210 | Improved |
| use ResFit instead of PA with test-time optimization | primitive fitting · 3DGen-Prim | #Prims: 64.87 -> 23.98 | Improved |
| add smooth union (ablation) | SuperFrustum · primitive ablation | IoU: 87.15 -> 88.37 | Improved |
| add smooth union (ablation) | SuperFrustum · primitive ablation | Overlap: 0.257 -> 0.199 | Improved |
| add taper and bend degrees of freedom (SuperFrustum vs SuperPrimitive) (ablation) | primitive · primitive ablation | IoU: 86.66 -> 87.15 | Improved |
| use SuperFrustum instead of superquadrics (ablation) | primitive · primitive ablation | IoU: 76.50 -> 87.15 | Improved |
| use SuperFrustum instead of cuboids (ablation) | primitive · primitive ablation | IoU: 82.33 -> 87.15 | Improved |
| add iterative residual fitting (ResFit vs single-shot) (ablation) | MSD init · fitting ablation | IoU: 87.95 -> 89.86 | Improved |
| add iterative residual fitting (ResFit vs single-shot) (ablation) | MSD init · fitting ablation | #Prims: 28.17 -> 23.46 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CAPRI-Net — CAPRI-Net: Learning Compact CAD Shapes With Adaptive Primitive Assembly (CVPR 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Unsupervised 3D Shape Reconstruction by Part Retrieval and Assembly (CVPR 2023 · similarity 0.87)
- Unified Primitive Proxies for Structured Shape Completion (CVPR 2026 · similarity 0.86)
- Supervised Fitting of Geometric Primitives to 3D Point Clouds (CVPR 2019 · similarity 0.86)
- DPF-Net: Combining Explicit Shape Priors in Deformable Primitive Field for Unsupervised Structural Reconstruction of 3D Objects (ICCV 2023 · similarity 0.86)
- Structure-Aware Surface Reconstruction via Primitive Assembly (ICCV 2023 · similarity 0.86)
- Differentiable Blocks World: Qualitative 3D Decomposition by Rendering Primitives (NeurIPS 2023 · similarity 0.86)
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.
