Dongyoung Kim1, Jinwoo Kim1, Junsang Yu2, Seon Joo Kim1
1Yonsei University, 2Samsung Advanced Institute of Technology
CVPR 2024
White balance (WB) algorithms in many commercial cameras assume single and uniform illumination, leading to undesirable results when multiple lighting sources with different chromaticities exist in the scene. Prior research on multi-illuminant WB typically predicts illumination at the pixel level without fully grasping the scene's actual lighting conditions, including the number and color of light sources. This often results in unnatural outcomes lacking in overall consistency. To handle this problem, we present a deep white balancing model that leverages the slot attention, where each slot is in charge of representing individual illuminants. This design enables the model to generate chromaticities and weight maps for individual illuminants, which are then fused to compose the final illumination map. Furthermore, we propose the centroid-matching loss, which regulates the activation of each slot based on the color range, thereby enhancing the model to separate illumination more effectively. Our method achieves the state-of-the-art performance on both single- and multi-illuminant WB benchmarks, and also offers additional information such as the number of illuminants in the scene and their chromaticity. This capability allows for illumination editing, an application not feasible with prior methods.
We propose novel centroid-matching loss
This loss guides whether slots are active or not based on the chromaticity of the GT illuminant
Without this loss, all slots would be activated and intervene in the prediction
But with the introduction of the centroid matching loss, the correct slots are activated appropriately and the rest are deactivated
Each scene in the demo video has two light sources, except for the first scene
Our model predicts the chromaticity and weight map for each light source
You can control the chromaticity of each light source using a fully decomposed illumination map
@inproceedings{kim2024attentive,
title={Attentive Illumination Decomposition Model for Multi-Illuminant White Balancing},
author={Kim, Dongyoung and Kim, Jinwoo and Yu, Junsang and Kim, Seon Joo},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
pages={25512--25521},
year={2024}
}
