Research map / Image generation / Image restoration
Non-Local Intrinsic Decomposition With Near-Infrared Priors
Ziang Cheng, Yinqiang Zheng, Shaodi You, Imari Sato
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 global reflectance energy (ablation) | RGB local · NIR-RGB testsetA | avg LMSE (x1000): 76.2 -> 62.5 | Improved |
| add global reflectance energy (ablation) | RGB local · NIR-RGB testsetA | avg MSE (x1000): 87.7 -> 57.0 | Improved |
| add NIR input and NIR shading prior (ablation) | RGB local · NIR-RGB testsetA | avg LMSE (x1000): 76.2 -> 58.4 | Improved |
| add NIR input and NIR shading prior (ablation) | RGB local · NIR-RGB testsetA | avg MSE (x1000): 87.7 -> 52.2 | Improved |
| add global reflectance energy (ablation) | NIR local · NIR-RGB testsetA | avg LMSE (x1000): 58.4 -> 53.4 | Improved |
| add global reflectance energy (ablation) | NIR local · NIR-RGB testsetA | avg MSE (x1000): 52.2 -> 40.5 | Improved |
| add NIR input and NIR shading prior (ablation) | RGB global · NIR-RGB testsetA | avg LMSE (x1000): 62.5 -> 53.4 | Improved |
| add NIR input and NIR shading prior (ablation) | RGB global · NIR-RGB testsetA | avg MSE (x1000): 57.0 -> 40.5 | Improved |
| use convolution-based global energy instead of clustering sparsity constraint | Gehler et al. · NIR-RGB testsetA | avg LMSE (x1000): 80.0 -> 62.5 | Improved |
| use convolution-based global energy instead of clustering sparsity constraint | Gehler et al. · NIR-RGB testsetA | avg MSE (x1000): 52.0 -> 57.0 | Hurt |
| relax i = s + r with a posterior term instead of vanilla Colour Retinex | Colour Retinex · NIR-RGB testsetA | avg LMSE (x1000): 144 -> 76.2 | Improved |
| use NIR-assisted energy instead of Colour Retinex | Colour Retinex · NIR-RGB testsetB | avg LMSE (x1000): 7.50 -> 7.03 | Improved |
Closest papers by meaning
- Unsupervised Learning for Intrinsic Image Decomposition From a Single Image (CVPR 2020 · similarity 0.91)
- A Survey on Intrinsic Images: Delving Deep into Lambert and Beyond (IJCV 2022 · similarity 0.90)
- VT-Intrinsic: Physics-Based Decomposition of Reflectance and Shading using a Single Visible-Thermal Image Pair (CVPR 2026 · similarity 0.89)
- Reflectance Adaptive Filtering Improves Intrinsic Image Estimation (CVPR 2017 · similarity 0.89)
- Unsupervised Intrinsic Image Decomposition Using Internal Self-Similarity Cues (TPAMI 2021 · similarity 0.88)
- IDArb: Intrinsic Decomposition for Arbitrary Number of Input Views and Illuminations (ICLR 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.
