CLCC: Contrastive Learning for Color Constancy
This work addresses color constancy for computer vision applications, offering incremental improvements in accuracy and robustness.
The paper tackles color constancy by proposing CLCC, a contrastive learning framework that uses raw-domain color augmentation to learn illuminant-dependent features, achieving a 17.5% relative improvement over a baseline on the NUS-8 dataset and reducing worst-case error by 28.7% in data-sparse regions.
In this paper, we present CLCC, a novel contrastive learning framework for color constancy. Contrastive learning has been applied for learning high-quality visual representations for image classification. One key aspect to yield useful representations for image classification is to design illuminant invariant augmentations. However, the illuminant invariant assumption conflicts with the nature of the color constancy task, which aims to estimate the illuminant given a raw image. Therefore, we construct effective contrastive pairs for learning better illuminant-dependent features via a novel raw-domain color augmentation. On the NUS-8 dataset, our method provides $17.5\%$ relative improvements over a strong baseline, reaching state-of-the-art performance without increasing model complexity. Furthermore, our method achieves competitive performance on the Gehler dataset with $3\times$ fewer parameters compared to top-ranking deep learning methods. More importantly, we show that our model is more robust to different scenes under close proximity of illuminants, significantly reducing $28.7\%$ worst-case error in data-sparse regions.