In this paper, we present a novel method for generating high-quality, stylized 3D avatars that utilizes pre-trained image-text diffusion models for data generation and a Generative Adversarial Network ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results