Monochromatic Image Colorization using Machine Learning

Abstract

The introduction of Artificial intelligence has opened doors to many automatic, unsupervised learning trends, which help to translate and acknowledge data. During the past years, the procedure of colorization of monochrome images has been greater or greater in several application fields, like restoration of old images or degraded images, and also, storage of monochrome images is more efficient when compared to colored images. This issue is not excessively presented because of an extremely high likelihood of conceivable outcomes during the designation of varied subtleties to the picture. A considerable lot of the new advancements in colorization have pictures with a normal format or exceptionally refined information, like semantic guides as the info. In the proposed system we are making use of Generative Adversarial Network (GAN). The final outcome is compared between the traditional deep neural network and the generative Model