Cross-language font style transfer

Abstract In this paper, we propose a cross-language font style transfer system that can synthesize new by observing only few samples from another language. Automatic synthesis is challenging task and has attracted much research interest. Most previous works addressed problem transferring the of given subset to content unseen ones. Nevertheless, they focused on in same many cases, need learn one language then apply it other languages. Existing methods make difficult accomplish because abstraction differences. To address problem, specifically designed network into multi-level attention form capture both local global features style. validate generative ability our model, constructed an experimental dataset 847 fonts, each containing English Chinese characters with Results show model generates 80.3% users’ preferred images compared state-of-the-art models.