Training a generative adversarial network (GAN) model thousands of image data derived from various acclaimed impressions of Wang Xizhi’s “Lanting Xu”, which enables the machine to generate its own dynamic recreation and interpolated impression of "Lanting".

​

由一个生成对抗网络在深度学习王羲之《兰亭集序》部分现存临本的图像数据，于上千字的多版兰亭数据中探寻某种规律、或然近于真笔的踪迹，而生成AI所书写的兰亭印象。

​

​

2020
Machine Learning, Image Data, Video, Projection Mapping