Dimma: Semi-supervised Low Light Image Enhancement with Adaptive Dimming

TL;DR

Dimma is a semi-supervised low-light image enhancement method that adapts to specific cameras using minimal paired data, employing a convolutional mixture density network and conditional UNet for high-quality, controllable results.

Contribution

We introduce a semi-supervised approach combining a convolutional mixture density network and conditional UNet, enabling effective low-light enhancement with minimal paired data and adjustable brightness control.

Findings

Achieves competitive results with few image pairs compared to fully supervised methods.

Surpasses state-of-the-art in some metrics when trained on full datasets.

Provides flexible brightness adjustment through user-controlled dimming factor.

Abstract

Enhancing low-light images while maintaining natural colors is a challenging problem due to camera processing variations and limited access to photos with ground-truth lighting conditions. The latter is a crucial factor for supervised methods that achieve good results on paired datasets but do not handle out-of-domain data well. On the other hand, unsupervised methods, while able to generalize, often yield lower-quality enhancements. To fill this gap, we propose Dimma, a semi-supervised approach that aligns with any camera by utilizing a small set of image pairs to replicate scenes captured under extreme lighting conditions taken by that specific camera. We achieve that by introducing a convolutional mixture density network that generates distorted colors of the scene based on the illumination differences. Additionally, our approach enables accurate grading of the dimming factor, which provides a wide range of control and flexibility in adjusting the brightness levels during the low-light image enhancement process. To further improve the quality of our results, we introduce an architecture based on a conditional UNet. The lightness value provided by the user serves as the conditional input to generate images with the desired lightness. Our approach using only few image pairs achieves competitive results compared to fully supervised methods. Moreover, when trained on the full dataset, our model surpasses state-of-the-art methods in some metrics and closely approaches them in others.