| Summary: Generative Adversarial Networks (GANs) are one of the state-of-the-art machine learning algorithms actively used today. Thanks to the capabilities of GANs, it has become possible to generate artificial images, sound or videos by processing complex examples. Models that generally use the GAN algorithm produce random results with random noise samples In this study, a new GAN model was designed. The model is about creating a result suitable for the conditions of the target face to be modified. The female,male exchange on the photographs given in the experiments was achieved with GAN in a remarkable way. The same model was trained in the Beardless,Bearded change, and it was determined that the model was generally successful. In addition, the success of the network to restore the photos to their original state was 90percentage.
Key Words: Generative Adversarial Networks (GANs), Machine Learning (ML). |
