StyleBank: An Explicit Representation for Neural Image Style Transfer

TL;DR

StyleBank introduces an explicit convolution filter bank approach for neural image style transfer, enabling flexible, incremental, and region-specific style transfer with real-time performance and improved quality.

Contribution

The paper presents StyleBank, a novel explicit style representation method that allows incremental learning and region-level style fusion, improving upon previous neural style transfer techniques.

Findings

Real-time style transfer with high-quality results

Supports incremental addition of new styles without retraining

Enables region-specific style fusion

Abstract

We propose StyleBank, which is composed of multiple convolution filter banks and each filter bank explicitly represents one style, for neural image style transfer. To transfer an image to a specific style, the corresponding filter bank is operated on top of the intermediate feature embedding produced by a single auto-encoder. The StyleBank and the auto-encoder are jointly learnt, where the learning is conducted in such a way that the auto-encoder does not encode any style information thanks to the flexibility introduced by the explicit filter bank representation. It also enables us to conduct incremental learning to add a new image style by learning a new filter bank while holding the auto-encoder fixed. The explicit style representation along with the flexible network design enables us to fuse styles at not only the image level, but also the region level. Our method is the first style transfer network that links back to traditional texton mapping methods, and hence provides new understanding on neural style transfer. Our method is easy to train, runs in real-time, and produces results that qualitatively better or at least comparable to existing methods.