Research map / 3D vision / Depth and stereo
Leveraging the Availability of Two Cameras for Illuminant Estimation
Abdelrahman Abdelhamed, Abhijith Punnappurath, Michael S. Brown
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 |
|---|---|---|---|
| Use a two-camera color transform as input to the illuminant estimator | Illuminant estimation on the radiometric dataset · Radiometric dataset | Mean angular error (degrees): 3.68 -> 2.65 | Improved |
| Use a two-camera color transform as input to the illuminant estimator | Illuminant estimation on the quasi-real NUS dataset · Quasi-real NUS dataset | Mean angular error (degrees): 2.04 -> 1.69 | Improved |
| Use a two-camera color transform as input to the illuminant estimator | Illuminant estimation on the S20 dataset · S20 two-camera dataset | Mean angular error (degrees): 2.14 -> 0.94 | Improved |
| Increase network size from 200 to 470 parameters (ablation) | Two-camera illuminant estimator · Radiometric dataset | Mean angular error (degrees): 2.80 -> 2.65 | Improved |
| Increase network size from 200 to 1460 parameters (ablation) | Two-camera illuminant estimator · Quasi-real NUS dataset | Mean angular error (degrees): 2.39 -> 1.69 | Improved |
| Increase network size from 470 to 1460 parameters (ablation) | Two-camera illuminant estimator · S20 two-camera dataset | Mean angular error (degrees): 0.94 -> 1.08 | Hurt |
| Add data augmentation by re-illuminating training image pairs (ablation) | Two-camera illuminant estimator · S20 real dataset | Mean angular error (degrees): 2.11 -> 1.73 | Improved |
| Add data augmentation by re-illuminating training image pairs (ablation) | Two-camera illuminant estimator · NUS dataset | Mean angular error (degrees): 4.74 -> 2.39 | Improved |
Closest papers by meaning
- A Multi-Hypothesis Approach to Color Constancy (CVPR 2020 · similarity 0.89)
- Neural Illumination: Lighting Prediction for Indoor Environments (CVPR 2019 · similarity 0.87)
- Two Illuminant Estimation and User Correction Preference (CVPR 2016 · similarity 0.87)
- Learning to Separate Multiple Illuminants in a Single Image (CVPR 2019 · similarity 0.87)
- Attentive Illumination Decomposition Model for Multi-Illuminant White Balancing (CVPR 2024 · similarity 0.85)
- LuxDiT: Lighting Estimation with Video Diffusion Transformer (NeurIPS 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.
