Research map / Image generation / Image restoration
Learning to Separate Multiple Illuminants in a Single Image
Zhuo Hui, Ayan Chakrabarti, Kalyan Sunkavalli, Aswin C. Sankaranarayanan
Check 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 intermediate chromaticity and shading supervision to the full cascade (ablation) | Illuminant-separation cascade · Synthetic benchmark | separated-image L1 error: 0.0351 -> 0.0288 | Improved |
| Use Hui et al.'s separation algorithm with the fully supervised model's chromaticity estimates instead of the network's direct outputs (ablation) | Fully supervised cascade · Synthetic benchmark | separated-image L1 error: 0.0288 -> 0.0207 | Improved |
| Train only ChromNet with chromaticity supervision rather than the full cascade (ablation) | Illuminant-separation cascade · Synthetic benchmark | separated-image L1 error: 0.0207 -> 0.0398 | Hurt |
| Train a single U-Net for direct separation instead of using the physically motivated cascade (ablation) | Illuminant separation · Synthetic benchmark | separated-image L1 error: 0.0351 -> 0.0679 | Hurt |
| Add shading and separated-image supervision alongside chromaticity supervision (ablation) | ChromNet · Synthetic benchmark | reflectance-chromaticity L1 error: 0.0308 -> 0.0537 | Hurt |
| Optimize chromaticity estimates for final separation quality rather than chromaticity accuracy alone (ablation) | Illuminant-separation cascade · Synthetic benchmark | separated-image L1 error: 0.0398 -> 0.0207 | Improved |
Closest papers by meaning
- Illuminant Spectra-Based Source Separation Using Flash Photography (CVPR 2018 · similarity 0.89)
- After the Party: Navigating the Mapping From Color to Ambient Lighting (ICCV 2025 · similarity 0.87)
- Leveraging the Availability of Two Cameras for Illuminant Estimation (CVPR 2021 · similarity 0.87)
- Attentive Illumination Decomposition Model for Multi-Illuminant White Balancing (CVPR 2024 · similarity 0.86)
- Flash-Split: 2D Reflection Removal with Flash Cues and Latent Diffusion Separation (CVPR 2025 · similarity 0.85)
- Diffusion Reflectance Map: Single-Image Stochastic Inverse Rendering of Illumination and Reflectance (CVPR 2024 · similarity 0.85)
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.
