A Dataset of Multi-Illumination Images in the Wild

TL;DR

This paper introduces a new dataset of over 1000 real scenes captured under 25 different lighting conditions to improve computer vision tasks involving lighting and material understanding.

Contribution

The paper presents a large, diverse multi-illumination dataset and demonstrates its usefulness by training models for illumination estimation, relighting, and white balance.

Findings

Models trained on the dataset outperform previous methods.

The dataset enables solving previously ill-posed lighting problems.

Demonstrates significant improvements in three applications.

Abstract

Collections of images under a single, uncontrolled illumination have enabled the rapid advancement of core computer vision tasks like classification, detection, and segmentation. But even with modern learning techniques, many inverse problems involving lighting and material understanding remain too severely ill-posed to be solved with single-illumination datasets. To fill this gap, we introduce a new multi-illumination dataset of more than 1000 real scenes, each captured under 25 lighting conditions. We demonstrate the richness of this dataset by training state-of-the-art models for three challenging applications: single-image illumination estimation, image relighting, and mixed-illuminant white balance.