Comfyui-CatVTON

Comfyui-CatVTON
★ 172

虚拟试衣扩散模型轻量化低显存推理
Comfyui-CatVTON节点，封装CatVTON虚拟试穿扩散模型，网络轻量、参数高效，可在低显存（<8GB）下生成1024×768试穿结果，兼容Detectron2/DensePose。
💡 在ComfyUI中快速生成1024×768虚拟试穿图像，节省显存并支持人体姿态配准
🍴 23 Forks💻 Jupyter Notebook🔄 2024-10-03
📦
网盘下载
复制链接后前往夸克网盘下载
https://pan.quark.cn/s/e58c8376a81b
📦 requirements.txt
opencv-python==4.4.0.46
img.png
📄 README

Comfyui-CatVTON: Concatenation Is All You Need for Virtual Try-On with Diffusion Models

Comfyui-CatVTON This repository is the modified official Comfyui node of CatVTON, which is a simple and efficient virtual try-on diffusion model with

*1) Lightweight Network (899.06M parameters totally)*,

*2) Parameter-Efficient Training (49.57M parameters trainable)*

*3) Simplified Inference (< 8G VRAM for 1024X768 resolution)*.

The original GitHub project is https://github.com/Zheng-Chong/CatVTON

Installation

  • git clone https://github.com/pzc163/Comfyui-CatVTON.git under the ComfyUI-aki-v1.3/custom_nodes path or install https://github.com/pzc163/Comfyui-CatVTON.git according to Comfyui Manager with git URL
  • install Detectron2 and DensePose
    • pip install git+https://github.com/facebookresearch/detectron2.git@v0.6
pip install git+https://github.com/facebookresearch/detectron2.git@v0.6#subdirectory=projects/DensePose

    For Windows OS users, if you encounter a compilation and installation failure prompt, you can choose between the following two options

    if python==3.10

    Please download Detectron2 and DensePose zip file in the Releases, which includes the code placed under /ComfyUI/python/Lib/site-packages of ComfyUI folder path.

    Notice that Detectron2 and DensePose zip file was compiled under python==3.10 Cuda==12.1 torch==2.1.2 environment if you can’t install Detectron2 and DensePose with the released zip file.

    if python==3.11

    Please download Detectron2 and DensePose whl file in the Releases, which includes the wheel file under python==3.11,placed under /ComfyUI/python/Lib/site-packages of ComfyUI folder path.

    then open cmd under ./ComfyUI/python/Lib/site-packages/ path

    pip install detectron2-0.6-cp311-cp311-win_amd64.whl

    pip install detectron2_densepose-0.6-py3-none-any.whl

  • Run the ComfyUI.
  • Download catvton_workflow.json and drag it into you ComfyUI webpage and enjoy 😆!

    • When you run the CatVTON workflow for the first time, the weight files will be automatically downloaded, which usually takes dozens of minutes.

    If you need to deploy catVTON in a anaconda environment, you can follow the steps below:

    An Installation Guide is provided to help build the conda environment for CatVTON. When deploying the app, you will need Detectron2 & DensePose, but these are not required for inference on datasets. Install the packages according to your needs.

    Reference

    Our code is modified based on https://github.com/Zheng-Chong/CatVTON