A Keras implementation with speed enhancements originally written in Tensorflow 1.x for Learning to See in the Dark by Chen Chen et. el. Original description can be found here: http://cchen156.web.engr.illinois.edu/SID.html. Original Paper: https://arxiv.org/abs/1805.01934
The original model was recreated it KERAS and was retrined with the dataset provided by the original authors, Chen Chen et. el for the Bayer filter (Sony camera dataset) and the results achived were at par with the results obtained by the authors as demonstrated in the paper.
Some of the results obtained are as below:
