Research map / 3D vision / Face and shape
Sketch2Mesh: Reconstructing and Editing 3D Shapes From Sketches
Benoit Guillard, Edoardo Remelli, Pierre Yvernay, Pascal Fua
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 |
|---|---|---|---|
| add Chamfer contour refinement (ablation) | MeshSDF encoder/decoder · Cars hand-drawn sketches (train Suggestive) | CD-l2 x1e3: 6.818 -> 4.395 | Improved |
| add Chamfer contour refinement (ablation) | MeshSDF encoder/decoder · Chairs hand-drawn sketches (train Suggestive) | CD-l2 x1e3: 18.752 -> 13.787 | Improved |
| use Chamfer refinement instead of rendering-based refinement (ablation) | Sketch2Mesh · Cars hand-drawn (train Suggestive) | CD-l2 x1e3: 5.752 -> 4.395 | Improved |
| use Chamfer refinement instead of rendering-based refinement (ablation) | Sketch2Mesh · Cars SketchFD (train Suggestive) | CD-l2 x1e3: 4.253 -> 3.132 | Improved |
| use Chamfer refinement instead of rendering-based refinement (ablation) | Sketch2Mesh · Chairs hand-drawn (train Suggestive) | CD-l2 x1e3: 17.519 -> 13.787 | Improved |
| use Chamfer refinement instead of rendering-based refinement (ablation) | Sketch2Mesh · Cars Suggestive (train Suggestive, in-style) | CD-l2 x1e3: 1.400 -> 1.420 | No clear effect |
| use Chamfer refinement instead of rendering-based refinement (ablation) | Sketch2Mesh · Chairs SketchFD (train SketchFD, in-style) | CD-l2 x1e3: 6.517 -> 6.737 | Hurt |
| add rendering-based refinement (ablation) | Sketch2Mesh · Cars Suggestive (train Suggestive) | CD-l2 x1e3: 1.613 -> 1.400 | Improved |
| feed predicted mask as extra network input | MeshSDF · Cars (train SketchFD, test Suggestive) | CD-l2 x1e3: 3.231 -> 3.124 | Improved |
| refine with Chamfer on contours instead of conditioning on mask | MeshSDF + mask · Cars (train SketchFD, test Suggestive) | CD-l2 x1e3: 3.124 -> 2.419 | Improved |
| replace MeshRCNN with Sketch2Mesh | comparison · Chairs Suggestive (train Suggestive) | CD-l2 x1e3: 6.775 -> 7.180 | Hurt |
| replace MeshRCNN with Sketch2Mesh | comparison · Chairs hand-drawn (train Suggestive) | CD-l2 x1e3: 19.055 -> 13.787 | Improved |
Later papers that compared with this method
1 later accepted paper reports Sketch2Mesh as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- PASTA: Part-Aware Sketch-to-3D Shape Generation with Text-Aligned Prior (ICCV 2025 · better on 6 of 6 shared results)
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.
- MeshSDF — MeshSDF: Differentiable Iso-Surface Extraction (NeurIPS 2020) · 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
- SketchSampler: Sketch-Based 3D Reconstruction via View-Dependent Depth Sampling (ECCV 2022 · similarity 0.90)
- Sketch2Model: View-Aware 3D Modeling From Single Free-Hand Sketches (CVPR 2021 · similarity 0.89)
- 3D Reconstruction From a Single Sketch via View-Dependent Depth Sampling (TPAMI 2024 · similarity 0.89)
- Doodle Your 3D: From Abstract Freehand Sketches to Precise 3D Shapes (CVPR 2024 · similarity 0.88)
- SketchINR: A First Look into Sketches as Implicit Neural Representations (CVPR 2024 · similarity 0.86)
- Diff3DS: Generating View-Consistent 3D Sketch via Differentiable Curve Rendering (ICLR 2025 · 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.
