Tero Karras, Samuli Laine, & Timo Aila (2019)
2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 4396-4405.
DOI: https://doi.org/10.1109/cvpr.2019.00453
Abstract. Introduces StyleGAN, whose mapping network and adaptive instance normalisation disentangle high-level attributes (pose, identity) from stochastic variation (freckles, hair) and enable unprecedented control over generated face images.
Tags: generative gan stylegan