Research map / 3D vision / Depth and stereo
Off the Planckian Locus: Using 2D Chromaticity to Improve In-Camera Color
SaiKiran Tedla, Joshua Little, Hakki Karaimer, 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 |
|---|---|---|---|
| use 2D chromaticity input instead of 1D CCT (ablation) | CCM-MLP · Lightbox dataset (Pixel) | mean angular error (deg): 4.11 -> 3.60 | Improved |
| use 2D chromaticity input instead of 1D CCT (ablation) | NN · Lightbox dataset (Pixel) | mean angular error (deg): 5.24 -> 4.30 | Improved |
| use 2D chromaticity input instead of 1D CCT (ablation) | CCCNN · Lightbox dataset (Pixel) | mean angular error (deg): 4.39 -> 4.00 | Improved |
| use 2D chromaticity input instead of 1D CCT (ablation) | EXPINV · Lightbox dataset (Pixel) | mean angular error (deg): 4.31 -> 3.84 | Improved |
| use 2D chromaticity input instead of 1D CCT (ablation) | CCM-MLP(RP) · Lightbox dataset (Pixel) | mean angular error (deg): 3.77 -> 3.27 | Improved |
| use xy chromaticity instead of raw 2D coordinates (ablation) | CCM-MLP · In-the-wild (Pixel) | mean angular error (deg): 4.02 -> 3.64 | Improved |
| add noise augmentation on input chromaticity (ablation) | CCM-MLP (2D xy) · In-the-wild (Pixel) | mean angular error (deg): 3.64 -> 3.32 | Improved |
| add noise augmentation on input chromaticity (ablation) | CCM-MLP (2D xy) · Lightbox dataset (Pixel) | mean angular error (deg): 3.53 -> 3.60 | Hurt |
| predict root-polynomial 3x6 mapping instead of linear 3x3 (ablation) | CCM-MLP (2D xy, noise aug) · Lightbox dataset (Pixel) | mean angular error (deg): 3.60 -> 3.27 | Improved |
| predict polynomial 3x9 mapping instead of linear 3x3 (ablation) | CCM-MLP (2D xy, noise aug) · Lightbox dataset (Pixel) | mean angular error (deg): 3.60 -> 3.67 | Hurt |
| use 2D chromaticity input instead of 1D CCT (ablation) | CCM-MLP · NUS (Nikon) | mean angular error (deg): 1.29 -> 1.28 | No clear effect |
| use 2D chromaticity input instead of 1D CCT (ablation) | CCM-MLP · LSMI single-illuminant (Galaxy) | mean angular error (deg): 3.77 -> 3.49 | Improved |
Closest papers by meaning
- A Multi-Hypothesis Approach to Color Constancy (CVPR 2020 · similarity 0.88)
- Improving Color Reproduction Accuracy on Cameras (CVPR 2018 · similarity 0.87)
- Luminance-GS: Adapting 3D Gaussian Splatting to Challenging Lighting Conditions with View-Adaptive Curve Adjustment (CVPR 2025 · similarity 0.84)
- Luminance-Aware Statistical Quantization: Unsupervised Hierarchical Learning for Illumination Enhancement (NeurIPS 2025 · similarity 0.84)
- GLUT: 3D Gaussian Lookup Table for Continuous Color Transformation (NeurIPS 2026 · similarity 0.84)
- Do Image Editing Models Understand Lighting? (NeurIPS 2026 · 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.
