A Retinex based GAN Pipeline to Utilize Paired and Unpaired Datasets for Enhancing Low Light Images
This addresses the problem of robust computer vision in low-light conditions, but it is incremental as it combines existing techniques into a new pipeline.
The paper tackles low-light image enhancement by introducing a deep learning pipeline that learns from both paired and unpaired datasets, achieving improved performance through the use of CNNs, GANs, cycle consistency loss, and a patched discriminator.
Low light image enhancement is an important challenge for the development of robust computer vision algorithms. The machine learning approaches to this have been either unsupervised, supervised based on paired dataset or supervised based on unpaired dataset. This paper presents a novel deep learning pipeline that can learn from both paired and unpaired datasets. Convolution Neural Networks (CNNs) that are optimized to minimize standard loss, and Generative Adversarial Networks (GANs) that are optimized to minimize the adversarial loss are used to achieve different steps of the low light image enhancement process. Cycle consistency loss and a patched discriminator are utilized to further improve the performance. The paper also analyses the functionality and the performance of different components, hidden layers, and the entire pipeline.