ComfyUI-DreamWaltz-G

★ 3

3D高斯人像骨骼引导扩散模型视频驱动

为ComfyUI提供与DreamWaltz-G集成的自定义节点，基于骨骼引导的2D扩散生成可动画的3D高斯人像，便于与视频驱动和姿态估计流水线结合。

💡 用自定义视频和TRAM姿态估计驱动生成动态3D头像。

🍴 1 Forks💻 Python🔄 2024-10-27

🔗 GitHub 原文

📦

网盘下载

复制链接后前往夸克网盘下载

https://pan.quark.cn/s/9d76119b2ef2

📦 requirements.txt

#git+https://github.com/nghorbani/body_visualizer
git+https://github.com/MPI-IS/configer.git
git+https://github.com/MPI-IS/mesh.git
tqdm
numpy
dotmap
pytorch-lightning
PyYAML
torch
transforms3d==0.3.1
pytorch3d
omegaconf
loguru

📄 README

ComfyUI-DreamWaltz-G

This repository contains custom ComfyUI nodes designed to integrate with DreamWaltz-G, a cutting-edge model for generating expressive 3D Gaussian avatars using skeleton-guided 2D diffusion.

Features (Planned)

DreamWaltz-G Nodes: Leverage DreamWaltz-G capabilities in ComfyUI.

3D Avatar Generation: Integrate 3D Gaussian avatars into your ComfyUI workflows.

Animate avatars: Using your own custom videos with SOTA 3D Human Pose estimation library TRAM)