Latest revision as of 23:02, 28 January 2021

Abstract

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.

Original document

The different versions of the original document can be found in:

• http://arxiv.org/abs/2006.15304

• http://arxiv.org/pdf/2006.15304

• http://xplorestaging.ieee.org/ielx7/9179991/9185188/09185373.pdf?arnumber=9185373,

http://dx.doi.org/10.1109/mercon50084.2020.9185373

• https://dblp.uni-trier.de/db/journals/corr/corr2006.html#abs-2006-15304,

https://arxiv.org/pdf/2006.15304,

https://ui.adsabs.harvard.edu/abs/2020arXiv200615304W/abstract,

https://academic.microsoft.com/#/detail/3083484956