GCC: Generative Color Constancy via Diffusing a Color Checker

TL;DR

GCC introduces a diffusion model-based method for color constancy that effectively generalizes across different camera sensors by inpainting color checkers and utilizing robust priors, without sensor-specific training.

Contribution

The paper proposes a novel diffusion model approach with inpainting, Laplacian decomposition, and data augmentation for improved cross-camera color constancy.

Findings

Strong robustness in cross-camera scenarios

Effective generalization without sensor-specific training

Utilizes pre-trained diffusion models for illumination estimation

Abstract

Color constancy methods often struggle to generalize across different camera sensors due to varying spectral sensitivities. We present GCC, which leverages diffusion models to inpaint color checkers into images for illumination estimation. Our key innovations include (1) a single-step deterministic inference approach that inpaints color checkers reflecting scene illumination, (2) a Laplacian decomposition technique that preserves checker structure while allowing illumination-dependent color adaptation, and (3) a mask-based data augmentation strategy for handling imprecise color checker annotations. By harnessing rich priors from pre-trained diffusion models, GCC demonstrates strong robustness in challenging cross-camera scenarios. These results highlight our method's effective generalization capability across different camera characteristics without requiring sensor-specific training, making it a versatile and practical solution for real-world applications.