Generative Models for Multi-Illumination Color Constancy
This addresses a domain-specific problem in computer vision for applications like photography and robotics, but it is incremental as it builds on existing GAN and color constancy techniques.
The paper tackles the problem of multi-illumination color constancy, where existing methods are limited to single light sources and datasets are scarce, by proposing a physics-driven method using GANs for image-to-image translation and a novel data augmentation technique, resulting in outperforming state-of-the-art methods on both single and multi-illumination datasets.
In this paper, the aim is multi-illumination color constancy. However, most of the existing color constancy methods are designed for single light sources. Furthermore, datasets for learning multiple illumination color constancy are largely missing. We propose a seed (physics driven) based multi-illumination color constancy method. GANs are exploited to model the illumination estimation problem as an image-to-image domain translation problem. Additionally, a novel multi-illumination data augmentation method is proposed. Experiments on single and multi-illumination datasets show that our methods outperform sota methods.