ProjectsProject Details

The goal of this work is to research the capability of generating a completely new image with the same visual content of a single given natural image, by using unsupervised learning of a deep neural network without the use of a GAN.

This project is based on the work presented in the paper: "SinGAN: Learning a Generative Model from a Single Natural Image" (Rot-Shahamet al.)
Different papers published in the last couple of years have already established the connection between the deep features of classification networks and the semantic content of images, such that we can define the visual content of an image by the statistics of its deep features. In the last couple of years, the rise of GANs for generating high-resolution images has also taken place. In the paper this project is based upon, a pyramid of GANs was trained, each responsible for a different scale of the image. This way the internal distribution of patches was learned, without any spatial dependency, and new images of arbitrary sizes could be generated. This method also allowed using GANs trained on a small training set (A single image).
Using a similar pyramidal structure, we succeeded in creating realistic and variable images in different sizes and aspects, that maintain both the global structure and the visual content of the source image, by using a single image, without using an adversarial loss. We also apply the method on several additional applications, taken from the image manipulation tasks of the original paper.