| （英） |
Image style transfer has drawn increasing attention recently. This task takes a content image and a style image as inputs and generates an output image with the desired style while preserving the structure from the content image. However, most existing approaches adopt a global feature transformation to transfer style patterns into content images (e.g., AdaIN and WCT). Such a design usually destroys the spatial information of the input images and fails to keep semantic consistency between the content image and the stylized results. To solve this problem, we propose a novel STyle TRansformer (STTR) network which breaks both content and style images into visual tokens to achieve a fine-grained style transformation. Specifically, two attention mechanisms are adopted in our STTR. We first propose to use self-attention to encode content and style tokens such that similar tokens can be grouped and learned together. We then adopt cross-attention between content and style tokens that encourages fine-grained style transformations. As similar patches lean to match with the same style code, semantic-consistent style transfer can be obtained. To compare STTR with existing approaches, we conduct user studies on Amazon Mechanical Turk (AMT), which are carried out with 50 human subjects with 1,000 votes in total. Extensive evaluations demonstrate the effectiveness and efficiency of the proposed STTR in generating visually pleasing style transfer results |