Abstract: While diffusion-based methods have shown impressive capabilities in capturing diverse and complex hairstyles, their ability to generate consistent and high-quality multi-view outputs—crucial ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results