Incorporating long-range consistency in CNN-based texture generation
This work addresses the limitation of capturing long-range consistency in texture generation for computer vision applications, though it appears incremental as it builds directly on an existing method.
The paper tackles the problem of generating images with long-range structure and symmetry constraints by modifying Gatys et al.'s CNN-based texture representation, resulting in improved rendering of regular textures and symmetric images, with applications to inpainting and season transfer.
Gatys et al. (2015) showed that pair-wise products of features in a convolutional network are a very effective representation of image textures. We propose a simple modification to that representation which makes it possible to incorporate long-range structure into image generation, and to render images that satisfy various symmetry constraints. We show how this can greatly improve rendering of regular textures and of images that contain other kinds of symmetric structure. We also present applications to inpainting and season transfer.