1 paper
Ryohei Suzuki, Masanori Koyama, Takeru Miyato +2
We present a novel CNN-based image editing strategy that allows the user to change the semantic information of an image over an arbitrary region by manipulating the feature-space r…