Sketch to image using CNN and GAN

Visual media has been one of the primary sources of human communication. We often have a propensity to rely on logos, images and other visual media to better represent, communicate, elucidate our thoughts. This proves vexing, as with any field that requires immense fecundity and originality in imagination. In this paper, we present a scintillating technique that alleviates human effort and time in creating new images. The technique contains essences of both Image Recognition and Image Generation. With the synergy of the models generation and recognition the system gains efficacy and a superior accuracy when compared to existing GANs. The system is able stimulate image of considerable resemblance to real images.