Research map / Image generation / Image restoration
Time-Aware Auto White Balance in Mobile Photography
Mahmoud Afifi, Luxi Zhao, Abhijith Punnappurath, Mohamed A. Abdelsalam, Ran Zhang, Michael S. Brown
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 time feature p, capture info and noise stats to histogram (ablation) | histogram-only model · validation set (neutral GT) | mean angular error: 2.03 -> 1.66 | Improved |
| add histogram feature to time and noise features (ablation) | time+noise model · validation set (neutral GT) | mean angular error: 2.37 -> 1.75 | Improved |
| add time feature p (ablation) | full model without p · validation set (neutral GT) | mean angular error: 1.89 -> 1.66 | Improved |
| add noise stats n and SNR stats r (ablation) | model with H, m, p · validation set (neutral GT) | mean angular error: 1.85 -> 1.66 | Improved |
| add noise stats n only (ablation) | model with H, m, p, r · validation set (neutral GT) | mean angular error: 1.72 -> 1.66 | No clear effect |
| add SNR stats r only (ablation) | model with H, m, p, n · validation set (neutral GT) | mean angular error: 1.67 -> 1.66 | No clear effect |
| add capture info m to time feature and noise (ablation) | H+p+n model · validation set (neutral GT) | mean angular error: 1.75 -> 1.66 | Improved |
| add histogram to complete time-capture feature (ablation) | time-capture model · indoor scenes | mean angular error: 3.05 -> 1.97 | Improved |
| add histogram to complete time-capture feature (ablation) | time-capture model · outdoor scenes | mean angular error: 2.12 -> 1.77 | Improved |
| add capture data to time feature only (ablation) | time-only model · indoor scenes | mean angular error: 9.39 -> 3.05 | Improved |
| add capture info (ISO, exposure) from scratch to FFCC | FFCC · test set (neutral GT) | mean angular error: 2.62 -> 2.31 | Improved |
| add SNR stats r | Ours on Simple Cube++ · Simple Cube++ | mean angular error: 1.08 -> 1.01 | Improved |
Closest papers by meaning
- Physically-Plausible Illumination Distribution Estimation (ICCV 2023 · similarity 0.88)
- Leveraging Multispectral Sensors for Color Correction in Mobile Cameras (CVPR 2026 · similarity 0.87)
- Attentive Illumination Decomposition Model for Multi-Illuminant White Balancing (CVPR 2024 · similarity 0.87)
- White-Balance First, Adjust Later: Cross-Camera Color Constancy via Vision-Language Evaluation (CVPR 2026 · similarity 0.84)
- GCC: Generative Color Constancy via Diffusing a Color Checker (CVPR 2025 · similarity 0.83)
- HVDualformer: Histogram-Vision Dual Transformer for White Balance (AAAI 2025 · 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.
