magic-research / magic-animate

[CVPR 2024] MagicAnimate: Temporally Consistent Human Image Animation using Diffusion Model
https://showlab.github.io/magicanimate/
BSD 3-Clause "New" or "Revised" License
10.5k stars 1.08k forks source link

Issues with IUV diagrams #143

Open TDIT-haha opened 10 months ago

TDIT-haha commented 10 months ago

Hello, I have reviewed this paper and think it is done very well. I have reviewed the issue of generating motion videos using dense poses in iusse and have also attempted to generate them myself. After trying it out on my own, I have a question: can this motion video be generated using a torso segmentation model? If possible, are there any requirements for the color of the mask of the segmented image?