Image Palette: Brushstroke Synthesis-based Style Transfer
Zheng Miao, Yan Zhang, Zhibin Zheng, Zhengxing Sun
Accepted to Pacific Graphics 2014 (PG 2014)
Painting style transfer is one kind technology where given the sample paintings with some specific art style, we can render the target paintings in the same style as the samples after some computation. In this paper we present a new approach to 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 paintings to get the same painting style as the samples. Based on the approach mentioned here, we also design and realize the corresponding painting system. The results show that the users can get a style-transferred personalized target painting just by providing a sample painting and least interactions.
PAPER: Image Palette: Brushstroke Synthesis-based Style Transfer
Overview
An overview of our approach. With input sample images and target images, the output result image is constructed with a two-stage process: Area-based stroke settings and Stroke-based transfer synthesis.
Results
Result of style transfer by our method. (a)Sample image. (b)Target image. (c)Result of style transfer from (a) to (b).(Please refer to the paper for clear image)
Results of style transfer from multiple samples. (a)Sample images. (b)Target image. (c) Result of style transfer from (a) to (b).
Comparison of results
(a)Sample image. (b)Target image. (c)Result in [LFB13]. (d)Result by our method.
which is consistent with the actual painting requirements,.
PAPER: Image Palette: Brushstroke Synthesis-based Style Transfer