Int-Papercut 剪纸风格图案智能生成系统
刘恩茂，柳丽娟，王俊武，金琦明，姚琤，应放天
剪纸是一种传统民间艺术，作为世界非物质文化遗产(ICH)，它表达了人们的社会生活和民间活动。但剪纸制作仍然依赖于丰富的图案创作经验和技巧，一定程度上限制了剪纸艺术的发展。基于这个局限，本文提出Int-Papercut，一个基于卷积神经网络(CNN)的剪纸图案生成系统，可以识别和标记输入照片的模式，并利用剪纸的基本符号匹配和填充，最后将照片转换成剪纸风格模式。两名剪纸专家和8名剪纸新手的定性实验数据表明，Int-Papercut降低了剪纸难度，提升了对剪纸和剪纸制作的理解。Int-Papercut支持用户使用自己喜爱的图片自由创作剪纸作品，将促进剪纸文化的保护、发展和应用。
Enmao Liu, Lijuan Liu, Junwu Wang, Qiming Jin, Cheng Yao, Fangtian Ying
Paper cutting is a traditional folk art, as a world intangible cultural heritage (ICH), it expresses people's social life and folk activities. However, papercut production still relies on rich pattern creation experience and fixed pattern, which limits the creation of paper-cut art, especially for novices. To address this gap, we put forward Int-Papercut, a new papercut pattern generation system based on convolutional Neural Network (CNN), which can recognize and mark the patterns of the input photos, and use the basic symbols of papercut to match and fill in, and finally convert the photos into clip-cut style patterns. Empirical qualitative data from two papercut experts and 8 papercut novices show Int-Papercut facilitated their understanding of paper cutting and production. Our system is expected to support users to freely create papercut works with their favorite pictures, and promote the protection, development and application of papercut culture.
已发表论文/Publications
