Research map / Point clouds / Shape and surface completion
ExtrudeNet: Unsupervised Inverse Sketch-and-Extrude for Shape Parsing
Daxuan Ren, Jianmin Zheng, Jianfei Cai, Jiatong Li, Junzhe 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 freeform profile curves instead of circle (ablation) | ExtrudeNet · ShapeNet Airplane | CD (x1e-3): 0.877 -> 0.66 | Improved |
| use freeform profile curves instead of polygon (ablation) | ExtrudeNet · ShapeNet Airplane | CD (x1e-3): 0.838 -> 0.66 | Improved |
| increase number of extruded shapes from 8 to 64 (ablation) | ExtrudeNet · ShapeNet Airplane | CD (x1e-3): 0.89 -> 0.66 | Improved |
| increase number of extruded shapes from 8 to 64 (ablation) | ExtrudeNet · ShapeNet Bench | CD (x1e-3): 1.41 -> 0.78 | Improved |
| increase number of extruded shapes from 32 to 64 (ablation) | ExtrudeNet · ShapeNet Chair | CD (x1e-3): 1.47 -> 1.47 | No clear effect |
| increase sample rate per curve from 80 to 400 (ablation) | ExtrudeNet · ShapeNet Airplane | CD: 0.725 -> 0.664 | Improved |
| increase curves per sketch from 2 to 4 (ablation) | ExtrudeNet · ShapeNet Airplane | CD: 0.754 -> 0.664 | Improved |
| increase curves per sketch from 4 to 8 (ablation) | ExtrudeNet · ShapeNet Airplane | CD: 0.664 -> 1.09 | Hurt |
| free sketch plane placement instead of side-only (ablation) | ExtrudeNet · ShapeNet Airplane | CD: 3.291 -> 0.664 | Improved |
| free sketch plane placement instead of top-only (ablation) | ExtrudeNet · ShapeNet Airplane | CD: 0.821 -> 0.664 | Improved |
| add 15% padding to testing point sampling (ablation) | ExtrudeNet · ShapeNet | artifacts | Improved |
| ExtrudeNet vs CSG-Stump | ExtrudeNet · ShapeNet 13 classes mean | CD (x1e-3): 1.32 -> 1.15 | Improved |
Later papers that compared with this method
1 later accepted paper reports ExtrudeNet as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Draw Step by Step: Reconstructing CAD Construction Sequences from Point Clouds via Multimodal Diffusion. (CVPR 2024 · 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
- MV2Cyl: Reconstructing 3D Extrusion Cylinders from Multi-View Images (NeurIPS 2024 · similarity 0.89)
- CAD-SIGNet: CAD Language Inference from Point Clouds using Layer-wise Sketch Instance Guided Attention (CVPR 2024 · similarity 0.87)
- Revisiting CAD Model Generation by Learning Raster Sketch (AAAI 2025 · similarity 0.86)
- From 2D CAD Drawings to 3D Parametric Models: A Vision-Language Approach (AAAI 2025 · similarity 0.84)
- MiCADangelo: Fine-Grained Reconstruction of Constrained CAD Models from 3D Scans (NeurIPS 2025 · similarity 0.84)
- PyTorchGeoNodes: Enabling Differentiable Shape Programs for 3D Shape Reconstruction (CVPR 2025 · similarity 0.84)
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.
