JointFontGAN: Joint Geometry-Content GAN for Font Generation via Few-Shot Learning
JointFontGAN: Joint Geometry-Content GAN for Font Generation via Few-Shot Learning
复制标题
DOI：
10.1145/3394171.3413705
复制
发表时间：
2020-10
期刊：
影响因子：
--
通讯作者：
Yankun Xi;Guoli Yan;Jing Hua;Z. Zhong
中科院分区：
文献类型：
--
作者：
Yankun Xi;Guoli Yan;Jing Hua;Z. Zhong
由于真实的世界中的字体和文本呈现出各种各样的视觉效果，因此在野外自动生成字体和文本设计是一项具有挑战性的任务。在本文中，我们提出了一种新的模型，JointFontGAN，派生字体，包括几何结构和形状内容的正确性和一致性与很少的字体样本。具体来说，我们设计了一种基于端到端深度学习的方法，通过新的多流扩展条件生成对抗网络（XcGAN）模型来生成字体，该模型同时联合学习和生成字体骨架和字体表示。它可以适应神经网络级别的几何可变性和内容可扩展性。然后，我们应用它，沿着与发达国家的高效和有效的一阶段模型，在字母和句子/段落的文本生成与标准和艺术/手写风格。大量的实验和比较表明，我们的方法优于国家的最先进的方法上收集的数据集，包括20 K字体（字母和标点符号）与不同的风格。
Automatic generation of font and text design in the wild is a challenging task since font and text in real world exhibit various visual effects. In this paper, we propose a novel model, JointFontGAN, to derive fonts, including both geometric structures and shape contents in correctness and consistency with very few font samples available. Specifically, we design an end-to-end deep learning based approach for font generation through the new multi-stream extended conditional generative adversarial network (XcGAN) models, which jointly learn and generate both font skeleton and glyph representations simultaneously. It can adapt to the geometric variability and content scalability at the neural network level. Then, we apply it, along with the developed efficient and effective one-stage model, to text generations in letters and sentences / paragraphs with both standard and artistic / handwriting styles. The extensive experiments and comparisons demonstrate that our approach outperforms the state-of-the-art methods on the collected datasets including 20K fonts (letters and punctuations) with different styles.