Incorporating long-range consistency in CNN-based texture generation

Guillaume Berger; Roland Memisevic

ICLR 2017poster59 citations

Incorporating long-range consistency in CNN-based texture generation

Abstract

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.

Computer visionDeep learning

BibTeX

@inproceedings{
berger2017incorporating,
title={Incorporating long-range consistency in {CNN}-based texture generation},
author={Guillaume Berger and Roland Memisevic},
booktitle={International Conference on Learning Representations},
year={2017},
url={https://openreview.net/forum?id=HyGTuv9eg}
}