Texturize a GAN Using a Single Image
This addresses the need for texture-specific image generation for users in creative or design fields, but it is incremental as it builds on existing GAN frameworks.
The paper tackles the problem of customizing a deep generative model to generate images matching the texture of a single reference image, such as a church, by adapting GANs with a method that modifies pre-trained weights using patch discriminator and laplacian adversarial losses, resulting in outputs that match texture while maintaining diversity and realism.
Can we customize a deep generative model which can generate images that can match the texture of some given image? When you see an image of a church, you may wonder if you can get similar pictures for that church. Here we present a method, for adapting GANs with one reference image, and then we can generate images that have similar textures to the given image. Specifically, we modify the weights of the pre-trained GAN model, guided by the reference image given by the user. We use a patch discriminator adversarial loss to encourage the output of the model to match the texture on the given image, also we use a laplacian adversarial loss to ensure diversity and realism, and alleviate the contradiction between the two losses. Experiments show that the proposed method can make the outputs of GANs match the texture of the given image as well as keep diversity and realism.