论文标题
HyperNST：神经风格转移的超网络
HyperNST: Hyper-Networks for Neural Style Transfer
论文作者
论文摘要
我们提出Hypernst；基于超网络和stylegan2架构的图像艺术风格的神经风格转移（NST）技术。我们的贡献是一种新的方法，用于诱导通过公制空间进行参数化的样式转移，并预先训练基于样式的视觉搜索（SBV）。我们首次展示了此类空间可用于驱动NST，从而从SBVS系统中启用了样式的应用程序和插值。技术贡献是一个超网络，可预测对型stylegan2在多种艺术内容（肖像）（肖像）上进行预训练的重量更新，可以使用面部区域的语义图在每个区域量身定制样式参数化。我们在保留良好的风格转移性能的同时，在内容保存方面显示了超过最高的内容。
We present HyperNST; a neural style transfer (NST) technique for the artistic stylization of images, based on Hyper-networks and the StyleGAN2 architecture. Our contribution is a novel method for inducing style transfer parameterized by a metric space, pre-trained for style-based visual search (SBVS). We show for the first time that such space may be used to drive NST, enabling the application and interpolation of styles from an SBVS system. The technical contribution is a hyper-network that predicts weight updates to a StyleGAN2 pre-trained over a diverse gamut of artistic content (portraits), tailoring the style parameterization on a per-region basis using a semantic map of the facial regions. We show HyperNST to exceed state of the art in content preservation for our stylized content while retaining good style transfer performance.
