Multiple Style Transfer via Variational AutoEncoder

Zhi-Song Liu
Vicky Kalogeiton
Marie-Paule Cani
GeoViC, LIX, École Polytechnique, CNRS, IP Paris

alt text


Modern works on style transfer focus on transferring style from a single image. Recently, some approaches study multiple style transfer; these, however, are either too slow or fail to mix multiple styles. We propose \net, a Variational AutoEncoder for latent space-based style transfer. It performs multiple style transfer by projecting nonlinear styles to a linear latent space, and it fuses different styles by linear interpolation and transfers the new style to the content image. To evaluate ST-VAE, we experiment on COCO for single and multiple style transfer. Moreover, we present a case study revealing that ST-VAE outperforms other methods while being faster, flexible, and setting a new path for multiple style transfer.

alt text




Multiple Style Transfer via Variational AutoEncoder
IEEE International Conference on Image Processing (ICIP), 2021


This work was supported supported by the the Google chair at École Polytechnique.