W-Net : One-Shot Arbitrary-StyleChinese Character Generationwith Deep Neural Networks

13 Dec 2018  ·  Haochuan Jiang, Guanyu Yang, Kaizhu Huang, and Rui ZHANG ·

Abstract. Due to the huge category number, the sophisticated com-binations of various strokes and radicals, and the free writing or print-ing styles, generating Chinese characters with diverse styles is alwaysconsidered as a difficult task. In this paper, an efficient and general-ized deep framework, namely, the W-Net, is introduced for the one-shotarbitrary-style Chinese character generation task. Specifically, given asingle character (one-shot) with a specific style (e.g., a printed font orhand-writing style), the proposed W-Net model is capable of learningand generating any arbitrary characters sharing the style similar to thegiven single character. Such appealing property was rarely seen in theliterature. We have compared the proposed W-Net framework to manyother competitive methods. Experimental results showed the proposedmethod is significantly superior in the one-shot setting.



  Add Datasets introduced or used in this paper

Results from the Paper

  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.