Research map / 3D vision / Depth and stereo
A Multi-Hypothesis Approach to Color Constancy
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Train one model using multiple device datasets instead of one model per device | Ours · NUS multi-device folds | median angular error: 1.48 -> 1.33 | Helped |
| Pretrain on NUS and Cube+ | Ours · Gehler-Shi | median angular error: 1.43 -> 1.32 | Helped |
| Pretrain on Gehler-Shi and Cube+ | Ours · NUS | median angular error: 1.61 -> 1.55 | Helped |
| Use 120 rather than 5 K-means candidates (ablation) | Ours · Gehler-Shi | median angular error: 2.06 -> 1.32 | Helped |
| Increase K-means candidates from 120 to 150 (ablation) | Ours · Gehler-Shi | median angular error: 1.32 -> 1.33 | No clear effect |
| Increase K-means candidates from 120 to 200 (ablation) | Ours · Gehler-Shi | median angular error: 1.32 -> 1.39 | Hurt |
| Use 120 rather than 5 K-means candidates (ablation) | Ours · Cube challenge, novel sensor | median angular error: 1.71 -> 1.31 | Helped |
| Increase K-means candidates from 120 to 150 (ablation) | Ours · Cube challenge, novel sensor | median angular error: 1.31 -> 1.32 | No clear effect |
| Replace K-means candidates with uniform sampling (ablation) | Ours · Cube challenge | median angular error: 1.06 -> 1.20 | Hurt |
| Replace K-means candidates with GMM sampling (ablation) | Ours · Cube challenge | median angular error: 1.06 -> 1.10 | Hurt |
Closest papers by meaning
- GCC: Generative Color Constancy via Diffusing a Color Checker (CVPR 2025 · similarity 0.90)
- FC4: Fully Convolutional Color Constancy With Confidence-Weighted Pooling (CVPR 2017 · similarity 0.90)
- White-Balance First, Adjust Later: Cross-Camera Color Constancy via Vision-Language Evaluation (CVPR 2026 · similarity 0.87)
- Attentive Illumination Decomposition Model for Multi-Illuminant White Balancing (CVPR 2024 · similarity 0.87)
- NightCC: Nighttime Color Constancy via Adaptive Channel Masking (CVPR 2024 · similarity 0.87)
- LuxDiT: Lighting Estimation with Video Diffusion Transformer (NeurIPS 2025 · 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.