Research map / 3D vision / Face and shape
Differentiable Convex Polyhedra Optimization from Multi-view Images
Daxuan Ren, Haiyi Mei, Hezi Shi, Jianmin Zheng, Jianfei Cai, Lei Yang
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 convex densification via Loop subdivision (ablation) | DiffConvex with 16 convexes · ablation shape | CD L2: 0.46 -> 0.32 | Improved |
| add random convex spawning (ablation) | DiffConvex with 16 convexes · ablation shape | CD L2: 0.54 -> 0.32 | Improved |
| increase number of convexes from 16 to 128 (ablation) | DiffConvex · ablation shape | CD L2: 0.32 -> 0.05 | Improved |
| increase number of convexes from 128 to 512 (ablation) | DiffConvex · ablation shape | CD L2: 0.05 -> 0.05 | No clear effect |
| image-supervised differentiable convex polyhedra instead of implicit-field-supervised CVXNet/BSPNet | ShapeNet reconstruction · ShapeNet mean | L1 Chamfer Distance: 0.024 (CVXNet) -> 0.022 | Improved |
| image-supervised convexes instead of CVXNet | ShapeNet reconstruction · ShapeNet mean | L2 Chamfer Distance x1000: 0.526 (CVXNet) -> 0.592 | Hurt |
| image-only supervision on categories with large cavities | DiffConvex · ShapeNet cabinet | L1 Chamfer Distance: 0.026 (CVXNet) -> 0.042 | Hurt |
| convex polyhedra with volumetric texture instead of superquadric blocks (DBW) | multiview reconstruction · NeRF Synthetic and DTU | L2 Chamfer Distance: Lego 0.91 -> 0.44; DTU S105 5.21 -> 2.62 | Improved |
| convex polyhedra instead of DBW | multiview reconstruction · DTU S24 | L2 Chamfer Distance: 3.25 -> 3.87 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- BSPNet — BSP-Net: Generating Compact Meshes via Binary Space Partitioning (CVPR 2020) · better on 4 of 4 shared results
- CVXNet — CvxNet: Learnable Convex Decomposition (CVPR 2020) · better on 2 of 4 shared results
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
- A General Differentiable Mesh Renderer for Image-Based 3D Reasoning (TPAMI 2020 · similarity 0.87)
- DMesh: A Differentiable Mesh Representation (NeurIPS 2024 · similarity 0.87)
- Shape Abstraction via Marching Differentiable Support Functions (CVPR 2025 · similarity 0.84)
- PolyhedronNet: Representation Learning for Polyhedra with Surface-attributed Graph (ICLR 2025 · similarity 0.84)
- LinPrim: Linear Primitives for Differentiable Volumetric Rendering (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.
