Manual colorization of black and white images is a laborious task and inefficient. It has been attempted using Photoshop editing, but it proves to be difficult as it requires extensive research and a picture can take up to one month to colorize. A pragmatic approach to the task is to implement sophisticated image colorization techniques. The literature on image colorization has been an area of interest in the last decade, as it stands at the confluence of two arcane disciplines, digital image processing and deep learning. Efforts have been made to use the ever-increasing accessibility of end-to-end deep learning models and leverage the benefits of transfer learning. Image features can be automatically extracted from the training data using deep learning models such as Convolutional Neural Networks (CNN). This can be expedited by human intervention and by using recently developed Generative Adversarial Networks (GAN). We implement image colorization using various CNN and GAN models while leveraging pre-trained models for better feature extraction and compare the performance of these models.