Animate Anyone
[Submitted on 12 Apr 2023 (v1), last revised 30 Oct 2023 (this version, v4)]
DreamPose: Fashion Image-to-Video Synthesis via Stable Diffusion
https://news.cnblogs.com/n/761605/
哈哈，挺有意思。
川大博士研究生，此人的github其他repo也不错。
Up to now, after 180,000 training sessions, this unofficial code implementation still seems unable to correctly learn information about the human skeleton. Sometimes, it even fails to generate a normal human figure, displaying only the background. Moreover, this background seems to resemble the style of the reference.
😄😄🚀🚀Due to the absence of official source code release, this unofficial code has not been thoroughly validated, and there are still many details to be verified. We welcome collaboration from the community to collectively implement and refine this algorithm！！！
https://github.com/guoqincode/Open-AnimateAnyone
This repository contains a simple and unofficial implementation of Animate Anyone. This project is built upon magic-animate and AnimateDiff. This implementation is first developed by Qin Guo and then assisted by Zhenzhi Wang.
This project is under continuous development in part-time, there may be bugs in the code, welcome to correct them, I will optimize the code after the pre-trained model is released!
In the current version, we recommend training on 8 or 16 A100,H100 (80G) at 512 or 768 resolution. Low resolution (256,384) does not give good results!!! (VAE is very poor at reconstruction at low resolution.)
