Generation of “Hand-Drawn” Images Using Deep Convolutional Generative Adversarial Networks

Abstract

Generative Adversarial Networks are nowadays being used for generating new and life-like images from datasets of old images. This paper studies the usefulness of a more improved algorithm called Deep Convoluted Generative Adversarial Networks for the generation of hand-drawn images using the case study of cartoon images. A dataset of cartoon images from a cartoon dataset is used for testing the model and the quality of the generated synthetic images are found to be good.