Unseen Style Transfer Based on a Conditional Fast Style Transfer Network

Keiji Yanai

Abstract:

In this paper, we propose a feed-forward neural style transfer network which can
transfer unseen arbitrary styles. To do that, first, we extend the fast neural style
transfer network proposed by Johnson et al. (2016) so that the network can learn
multiple styles at the same time by adding a conditional input. We call this as “a
conditional style transfer network”. Next, we add a style condition network which
generates a conditional signal from a style image directly, and train “a conditional
style transfer network with a style condition network” in an end-to-end manner.
The proposed network can generate a stylized image from a content image and a
style image in one-time feed-forward computation instantly.

TL;DR: A feed-forward neural style transfer network which can transfer unseen arbitrary styles

Conflicts: uec.ac.jp