Single and Multiple Illuminant Estimation Using Convolutional Neural Networks

TL;DR

This paper introduces a CNN-based approach for estimating the color of illuminants in RAW images, capable of handling both single and multiple light sources with improved accuracy over existing methods.

Contribution

It presents a novel CNN architecture that produces local illuminant estimates and a detector to decide when to aggregate these estimates for multiple illuminants.

Findings

Lower estimation errors compared to state-of-the-art methods

Effective handling of both single and multiple illuminant scenarios

Demonstrated on standard datasets with improved accuracy

Abstract

In this paper we present a method for the estimation of the color of the illuminant in RAW images. The method includes a Convolutional Neural Network that has been specially designed to produce multiple local estimates. A multiple illuminant detector determines whether or not the local outputs of the network must be aggregated into a single estimate. We evaluated our method on standard datasets with single and multiple illuminants, obtaining lower estimation errors with respect to those obtained by other general purpose methods in the state of the art.