Image Palette: Brushstroke Synthesis-based Style Transfer

Abstract

Painting style transfer is one kind technology where given the sample images with some specific art style, we can render the target images in the same style as the samples after some computation. In this paper we present a new approach of painting style transfer in which such a style transfer work is done by analogy with simulating the process of creation. We take the sample as palette where users can select arbitrary outlines or textures as the current input mode brush strokes. We then analyze brush strokes' style feature information and use such information for the style transfer and synthesis along the stroke curves learned from the specified area in target images to get the same painting style as the samples. The results show that the users can get a style-transferred personalized target image just by the given sample images and least interactions.