Research map / 3D vision / Depth and stereo
Optimizing Illuminant Estimation in Dual-Exposure HDR Imaging
Mahmoud Afifi, Zhenhua Hu, Liang Liang
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 covariance matrix of the ratio image to DEF (ablation) | EMLP · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 4.07 -> 3.77 | Improved |
| use DEF to generate the bias map in CCC (ablation) | ECCC · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 4.02 -> 3.56 | Improved |
| add bias initialization from clustered ground-truth illuminants (ablation) | ECCC · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 3.99 -> 3.56 | Improved |
| use both long and short exposure histograms instead of a single or averaged one (ablation) | ECCC · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 4.03 -> 3.56 | Improved |
| increase exposure factor e from 2 to 8 (ablation) | EMLP · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 4.35 -> 4.07 | Improved |
| increase exposure factor e for ECCC (ablation) | ECCC · Dual-exposure HDR illuminant dataset (test set) | mean angular error: 2.94 (e=2), 3.18 (e=4), 3.00 (e=8) | Mixed |
| add chromatic-adaptation data augmentation (ablation) | EMLP · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 3.77 -> 3.52 | Improved |
| add chromatic-adaptation data augmentation (ablation) | ECCC · Dual-exposure HDR illuminant dataset (test set) | mean angular error: 3.00 -> 2.95 | No clear effect |
| use 3x3 color matrix instead of affine or homography mapping (ablation) | EMLP · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 4.78 -> 4.07 | Improved |
| ensemble EMLP and ECCC by averaging predicted illuminants | EMLP + ECCC · Dual-exposure HDR illuminant dataset (test set) | mean angular error: 3.00 -> 2.91 | Improved |
| feed both exposure histograms to C5 hypernetwork instead of an averaged one | C5 · Dual-exposure HDR illuminant dataset (validation set) | mean angular error: 3.93 -> 4.49 | Hurt |
Closest papers by meaning
- Color Event Enhanced Single-Exposure HDR Imaging (AAAI 2024 · similarity 0.89)
- Exposure-Aware Dynamic Weighted Learning for Single-Shot HDR Imaging (ECCV 2022 · similarity 0.88)
- Joint Demosaicing and Deghosting of Time-Varying Exposures for Single-Shot HDR Imaging (ICCV 2023 · similarity 0.86)
- High Dynamic Range Novel View Synthesis with Single Exposure (ICML 2025 · similarity 0.86)
- Self-Supervised High Dynamic Range Imaging with Multi-Exposure Images in Dynamic Scenes (ICLR 2024 · similarity 0.86)
- HDR-GS: Efficient High Dynamic Range Novel View Synthesis at 1000x Speed via Gaussian Splatting (NeurIPS 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.
