Research map / 3D vision / Face and shape
Tree-Structured Shading Decomposition
Chen Geng, Hong-Xing Yu, Sharon Zhang, Maneesh Agrawala, Jiajun Wu
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 two-stage tree decomposition (amortized autoregressive stage plus optimization) instead of CNN/LSTM/Transformer baselines | shade tree decomposition · Visual Shading Editing Analogy | PSNR: 5.02 -> 32.17 (vs Transformer) | Improved |
| add multiple sampling at first-stage inference (ablation) | tree decomposition method · shading decomposition | reconstruction quality (qualitative) | Improved |
| add second-stage structure optimization (ablation) | tree decomposition method · shading decomposition | reconstruction quality (qualitative) | Improved |
| add overall optimization at end of second stage (ablation) | tree decomposition method · shading decomposition | reconstruction quality (qualitative) | Improved |
Closest papers by meaning
- An Approximate Shading Model with Detail Decomposition for Object Relighting (IJCV 2018 · similarity 0.88)
- Shading Annotations in the Wild (CVPR 2017 · similarity 0.86)
- MaterialRefGS: Reflective Gaussian Splatting with Multi-view Consistent Material Inference (NeurIPS 2025 · similarity 0.85)
- RNG: Relightable Neural Gaussians (CVPR 2025 · similarity 0.84)
- Eclipse: Disambiguating Illumination and Materials using Unintended Shadows (CVPR 2024 · similarity 0.84)
- COCOTree: A Dataset and Benchmark for Open Tree-Structured Visual Decomposition (NeurIPS 2026 · 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.
