Research map / Image generation / Image restoration
GLoSH: Global-Local Spherical Harmonics for Intrinsic Image Decomposition
Hao Zhou, Xiang Yu, David W. Jacobs
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 synthetic SUNCG pre-training (ablation) | GLoSH · IIW | WHDR (%): 17.82 -> 15.20 | Improved |
| add synthetic SUNCG pre-training (ablation) | GLoSH · NYUv2 | mean angular error: 35.14 -> 25.57 | Improved |
| add synthetic SUNCG pre-training (ablation) | GLoSH · SAW | AP (%): 88.52 -> 95.01 | Improved |
| add self-supervised pseudo-label loss (ablation) | GLoSH · IIW | WHDR (%): 15.50 -> 15.20 | Improved |
| add self-supervised pseudo-label loss (ablation) | GLoSH · NYUv2 | mean angular error: 25.93 -> 25.57 | Improved |
| add self-supervised pseudo-label loss (ablation) | GLoSH · SAW | AP (%): 95.79 -> 95.01 | Hurt |
| add symmetric WHDR loss L_R2 (ablation) | GLoSH · SAW | AP (%): 91.89 -> 95.01 | Improved |
| add symmetric WHDR loss L_R2 (ablation) | GLoSH · IIW | WHDR (%): 15.34 -> 15.20 | Improved |
| add second fine scale to one-scale model (ablation) | GLoSH · IIW | WHDR (%): 18.70 -> 16.62 | Improved |
| add third scale to two-scale model (ablation) | GLoSH · IIW | WHDR (%): 16.62 -> 15.20 | Improved |
| add non-negative lighting constraint (ablation) | GLoSH coarse net · SUNCG val | negative-lighting global SH (%): 13.39% -> 1.09% | Improved |
| add global-local SH with coarse-to-fine net vs RGB-D baseline | GLoSH · NYUv2 shading | AP (%): 90.38% -> 95.43% | Improved |
Later papers that compared with this method
1 later accepted paper reports GLoSH as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Erasing Appearance Preservation in Optimization-based Smoothing (ECCV 2020 · better on 2 of 2 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
- Self-Supervised Intrinsic Image Decomposition (NeurIPS 2017 · similarity 0.86)
- ShadingNet: Image Intrinsics by Fine-Grained Shading Decomposition (IJCV 2021 · similarity 0.86)
- VT-Intrinsic: Physics-Based Decomposition of Reflectance and Shading using a Single Visible-Thermal Image Pair (CVPR 2026 · similarity 0.86)
- A Survey on Intrinsic Images: Delving Deep into Lambert and Beyond (IJCV 2022 · similarity 0.86)
- SpecNeRF: Gaussian Directional Encoding for Specular Reflections (CVPR 2024 · similarity 0.85)
- MetaGS: A Meta-Learned Gaussian-Phong Model for Out-of-Distribution 3D Scene Relighting (NeurIPS 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.
