Research map / 3D vision / Face and shape
Differentiable Stereopsis: Meshes From Multiple Views Using Differentiable Rendering
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 texture transfer instead of optimizing a fixed-UV texture image (ablation) | Differentiable Stereopsis · GSO, 8 views, 30° camera noise | Chamfer: 0.22 -> 0.10 | Helped |
| use texture transfer instead of optimizing a fixed-UV texture image (ablation) | Differentiable Stereopsis · GSO, 8 views, 30° camera noise | F1@.1: 29.5 -> 63.5 | Helped |
| use texture transfer instead of optimizing a fixed-UV texture image (ablation) | Differentiable Stereopsis · GSO, 8 views, 30° camera noise | F1@.2: 53.0 -> 80.6 | Helped |
| add texture information to the optimization (ablation) | Differentiable Stereopsis · GSO, 8 views, 20° camera noise | shape concavity detail | Helped |
| use texture transfer rather than naive UV texture optimization (ablation) | Differentiable Stereopsis · GSO, 8 views, 20° camera noise | shape and camera quality | Helped |
| use DS rather than NeRF-opt | 3D reconstruction · Tanks and Temples, Caterpillar, 15 views | Chamfer: 0.15 -> 0.07 | Helped |
| use DS rather than NeRF-opt | 3D reconstruction · Tanks and Temples, Barn, 15 views | Chamfer: 0.98 -> 0.39 | Hurt |
Closest papers by meaning
- Differentiable Volumetric Rendering: Learning Implicit 3D Representations Without 3D Supervision (CVPR 2020 · similarity 0.91)
- Learning a Multi-View Stereo Machine (NeurIPS 2017 · similarity 0.90)
- Extracting Triangular 3D Models, Materials, and Lighting From Images (CVPR 2022 · similarity 0.88)
- GTR: Improving Large 3D Reconstruction Models through Geometry and Texture Refinement (ICLR 2025 · similarity 0.86)
- Differentiable Display Photometric Stereo (CVPR 2024 · similarity 0.86)
- LinPrim: Linear Primitives for Differentiable Volumetric Rendering (NeurIPS 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.