Giters

Francis-Rings / MotionEditor

[CVPR2024] MotionEditor is the first diffusion-based model capable of video motion editing.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Dataset

AsukaCamellia opened this issue · comments

commented

Thanks for your wonderful work. Where is the preprocessing of the dataset?

commented

Thanks for your attention! We implement SAM and OpenPose to exact protagonist masks and pose skeletons of the given videos, respectively.

commented

Thank you for your excellent work. Where can I download the dataset used in the project?

commented

Thanks for your attention! The codes of data preparation are available to the public. MotionEditor is a one-shot learning-based diffusion model. Consequently, MotionEditor doesn't need any training datasets. If you would like to access the source videos utilized for inference, please feel free to send me an email for further details.