Research map / 3D vision / Face and shape
Extracting Triangular 3D Models, Materials, and Lighting From Images
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 explicit mesh optimization with factorized materials and lighting instead of NeRFactor | NeRFactor synthetic relighting · NeRFactor synthetic, 4-scene average | PSNR: 23.78 -> 24.53 | Helped |
| Use explicit mesh optimization with factorized materials and lighting instead of NeRFactor | NeRFactor synthetic relighting · NeRFactor synthetic, 4-scene average | LPIPS: 0.112 -> 0.085 | Helped |
| Use explicit mesh optimization rather than Mip-NeRF | NeRF realistic synthetic view interpolation · NeRF realistic synthetic, 8-scene average | PSNR: 33.05 -> 29.05 | Hurt |
| Use explicit mesh optimization rather than Mip-NeRF | NeRF realistic synthetic view interpolation · NeRF realistic synthetic, 8-scene average | LPIPS: 0.067 -> 0.081 | Hurt |
| Use explicit mesh optimization with factorization rather than NeRF | NeRFactor synthetic view interpolation · NeRFactor synthetic, 4-scene average | PSNR: 31.08 -> 31.65 | Helped |
| Use explicit mesh optimization instead of NeuS | DTU MVS reconstruction · scan65 | Chamfer loss: 0.72 -> 1.03 | Hurt |
| Use explicit mesh optimization instead of NeuS | DTU MVS reconstruction · scan106 | Chamfer loss: 0.66 -> 1.07 | Hurt |
| Use explicit mesh optimization instead of NeuS | DTU MVS reconstruction · scan118 | Chamfer loss: 0.51 -> 0.69 | Hurt |
| Use the proposed SDF sign-flip regularizer rather than no SDF regularization (ablation) | DMTet geometry · multi-view reconstruction | internal geometry and floaters | Helped |
| Use the proposed SDF regularizer rather than L1 SDF smoothness (ablation) | DMTet geometry · multi-view reconstruction | internal geometry removal | Helped |
| Add visibility-based regularization for faces unseen in training views (ablation) | DMTet geometry · DTU scan65 | floaters behind the visible surface | Helped |
| Parameterize DMTet SDF with an MLP instead of direct per-vertex values (ablation) | DMTet on ambiguous geometry · DTU MVS scans | surface smoothness | Helped |
Closest papers by meaning
- ReLi3D: Relightable Multi-view 3D Reconstruction with Disentangled Illumination (ICLR 2026 · similarity 0.91)
- QUADify: Extracting Meshes with Pixel-level Details and Materials from Images (CVPR 2024 · similarity 0.90)
- Eclipse: Disambiguating Illumination and Materials using Unintended Shadows (CVPR 2024 · similarity 0.89)
- IDArb: Intrinsic Decomposition for Arbitrary Number of Input Views and Illuminations (ICLR 2025 · similarity 0.89)
- Material Anything: Generating Materials for Any 3D Object via Diffusion (CVPR 2025 · similarity 0.87)
- Material Palette: Extraction of Materials from a Single Image (CVPR 2024 · similarity 0.87)
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.
