Aim of the project is to create a model that takes a grayscale (black and white) image as input and then produces a visually plausible and perceptually meaningful colorized output image as output. We used U-net and conditional GAN to solve the problem at hand. To avoid the problem of the blind leading the blind in the GAN game where neither generator nor discriminator knows anything about the task at the beginning of training, we pretrained the generator separately. Currently, it is able to colorize the input images pretty well with some flaws in colorizing blue colour. We took frames from a black and white video and tried colorizing it, the results were consistent.