Research map / Image generation / Image restoration
Multispectral Illumination Estimation Using Deep Unrolling Network
Yuqi Li, Qiang Fu, Wolfgang Heidrich
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 |
|---|---|---|---|
| make unrolled-stage parameters trainable instead of fixed values (ablation) | Unrolling illumination estimation network · Own dataset, multiple illumination | absolute error dS: 4.23 -> 2.84 | Improved |
| add TV regularization block (ablation) | Unrolling illumination estimation network · Own dataset, multiple illumination | absolute error dS: 4.35 -> 2.84 | Improved |
| add low-rank regularization block (ablation) | Unrolling illumination estimation network · Own dataset, multiple illumination | absolute error dS: 4.21 -> 2.84 | Improved |
| add non-local self-attention transformer block (ablation) | Unrolling illumination estimation network · Own dataset, multiple illumination | absolute error dS: 3.63 -> 2.84 | Improved |
| add non-local self-attention transformer block (ablation) | Unrolling illumination estimation network · Own dataset, multiple illumination | angular error dA: 0.25 -> 0.21 | Improved |
| use transformer instead of residual conv block for reflectance denoising (ablation) | Unrolling illumination estimation network · Own dataset, multiple illumination | absolute error dS: 3.50 -> 2.84 | Improved |
| use unrolling network instead of GrayEdge (best baseline) | Unrolling network · Own dataset, single global illumination | absolute error dS: 2.29 -> 1.80 | Improved |
| use unrolling network instead of PWIR (CNN) on unseen CAVE dataset | Unrolling network · CAVE, single global illumination | absolute error dS: 4.62 -> 1.87 | Improved |
| use unrolling network instead of AngularGAN | Unrolling network · Own dataset, multiple illumination | absolute error dS: 3.77 -> 2.84 | Improved |
| use unrolling network instead of BUTD | Unrolling network · CAVE, multiple illumination | angular error dA: 0.48 -> 0.31 | Improved |
Closest papers by meaning
- Illumination Spectrum Estimation for Multispectral Images via Surface Reflectance Modeling and Spatial-Spectral Feature Generation (CVPR 2025 · similarity 0.90)
- A Weighted Variational Model for Simultaneous Reflectance and Illumination Estimation (CVPR 2016 · similarity 0.86)
- Spectral Compressive Imaging via Chromaticity-Intensity Decomposition (NeurIPS 2025 · similarity 0.85)
- Learning to Separate Multiple Illuminants in a Single Image (CVPR 2019 · similarity 0.85)
- Diffusion Reflectance Map: Single-Image Stochastic Inverse Rendering of Illumination and Reflectance (CVPR 2024 · similarity 0.84)
- SPECAT: SPatial-spEctral Cumulative-Attention Transformer for High-Resolution Hyperspectral Image Reconstruction (CVPR 2024 · similarity 0.83)
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.
