1 paper
Zhiling Ye, Cong Zhou, Xiubao Zhang +3
In this paper, we explore a reconstruction and reenactment separated framework for 3D Gaussians head, which requires only a single portrait image as input to generate controllable…