P2-GAN: Efficient Stroke Style Transfer Using Single Style Image

Style transfer is a useful image synthesis technique that can re-render given image into another artistic style while preserving its content information. Generative Adversarial Network (GAN) is a widely adopted framework toward this task for its better representation ability on local style patterns than the traditional Gram-matrix based methods. However, most previous methods rely on sufficient amount of pre-collected style images to train the model. In this paper, a novel Patch Permutation GAN (P-2-GAN) network that can efficiently learn the stroke style from a single style image is proposed. We use patch permutation to generate multiple training samples from the given style image. A patch discriminator that can simultaneously process patch-wise images and natural images seamlessly is designed. We also propose a local texture descriptor based criterion to quantitatively evaluate the style transfer quality. Experimental results showed that our method can produce finer quality re-renderings from single style image with improved computational efficiency compared with many state-of-the-arts methods.

Automated summary

This paper proposes a novel network method called Patch Permutation GAN (P-2-GAN) that can learn stroke style from a single style image efficiently. Patch permutation is used to generate multiple training samples, and a patch discriminator that can process both patch-wise images and natural images is designed. A local texture descriptor based criterion is also proposed to evaluate the style transfer quality quantitatively. Experimental results demonstrate that our method can produce finer quality re-renderings with improved computational efficiency compared to state-of-the-art methods.