A Switched View of Retinex: Deep Self-Regularized Low-Light Image Enhancement

TL;DR

This paper introduces a novel self-regularized low-light image enhancement method based on Retinex theory that preserves colors and improves generalization across lighting conditions without needing paired training data.

Contribution

It proposes a Retinex-inspired approach that decouples color and brightness, using a reflectance estimation network with consistency constraints for effective enhancement.

Findings

Outperforms state-of-the-art algorithms qualitatively and quantitatively.

Effectively preserves colors and adapts to various lighting conditions.

Efficiently decouples image into color and brightness for better enhancement.

Abstract

Self-regularized low-light image enhancement does not require any normal-light image in training, thereby freeing from the chains on paired or unpaired low-/normal-images. However, existing methods suffer color deviation and fail to generalize to various lighting conditions. This paper presents a novel self-regularized method based on Retinex, which, inspired by HSV, preserves all colors (Hue, Saturation) and only integrates Retinex theory into brightness (Value). We build a reflectance estimation network by restricting the consistency of reflectances embedded in both the original and a novel random disturbed form of the brightness of the same scene. The generated reflectance, which is assumed to be irrelevant of illumination by Retinex, is treated as enhanced brightness. Our method is efficient as a low-light image is decoupled into two subspaces, color and brightness, for better preservation and enhancement. Extensive experiments demonstrate that our method outperforms multiple state-of-the-art algorithms qualitatively and quantitatively and adapts to more lighting conditions.