Texture Networks: Feed-forward Synthesis of Textures and Stylized Images

TL;DR

This paper introduces a fast, feed-forward neural network approach for texture synthesis and style transfer, achieving comparable quality to previous optimization-based methods but with significantly improved speed and efficiency.

Contribution

It presents a novel training method for convolutional networks that enables real-time texture synthesis and style transfer from a single example, reducing computational costs.

Findings

Networks generate high-quality textures comparable to optimization methods.

The approach is hundreds of times faster than previous techniques.

The method is flexible and can transfer artistic styles effectively.

Abstract

Gatys et al. recently demonstrated that deep networks can generate beautiful textures and stylized images from a single texture example. However, their methods requires a slow and memory-consuming optimization process. We propose here an alternative approach that moves the computational burden to a learning stage. Given a single example of a texture, our approach trains compact feed-forward convolutional networks to generate multiple samples of the same texture of arbitrary size and to transfer artistic style from a given image to any other image. The resulting networks are remarkably light-weight and can generate textures of quality comparable to Gatys~et~al., but hundreds of times faster. More generally, our approach highlights the power and flexibility of generative feed-forward models trained with complex and expressive loss functions.