ComfyUI Extension: Comfyui-MusePose

Authored by TMElyralab

Created

Updated

429 stars

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

a/MusePose is an image-to-video generation framework for virtual human under control signal such as pose. NOTE: You need to download weigths manually from: a/https://huggingface.co/TMElyralab/MusePose.[w/The repository name has changed. If you are not receiving updates, please delete the existing node and reinstall it.]

Looking for a different extension?

Custom Nodes (0)

    README

    MusePose

    MusePose is an image-to-video generation framework for virtual human under control signal such as pose.

    MusePose is the last building block of the Muse opensource serie. Together with MuseV and MuseTalk, we hope the community can join us and march towards the vision where a virtual human can be generated end2end with native ability of full body movement and interaction. Please stay tuned for our next milestone!

    Comfyui-MusePose

    If you're running on Linux, or non-admin account on windows you'll want to ensure /ComfyUI/custom_nodes and Comfyui-MusePose has write permissions.

    Followed ComfyUI's manual installation steps and do the following:

    • Navigate to your /ComfyUI/custom_nodes/ folder
    • Run git clone https://github.com/TMElyralab/Comfyui-MusePose.git
    • Navigate to your /ComfyUI/custom_nodes/Comfyui-MusePose folder and run
     pip install -r requirements.txt
    
     pip install --no-cache-dir -U openmim 
     mim install mmengine 
     mim install "mmcv>=2.0.1" 
     mim install "mmdet>=3.1.0" 
     mim install "mmpose>=1.1.0" 
    
    • Start ComfyUI

    Updates

    • requirements.txt: diffusers 0.27.2 is now supported

    Download weights

    You can download weights manually as follows:

    1. Download our trained weights.

    2. Download the weights of other components:

    Finally, these weights should be organized in pretrained_weights as follows:

    ./pretrained_weights/
    |-- MusePose
    |   |-- denoising_unet.pth
    |   |-- motion_module.pth
    |   |-- pose_guider.pth
    |   └── reference_unet.pth
    |-- dwpose
    |   |-- dw-ll_ucoco_384.pth
    |   └── yolox_l_8x8_300e_coco.pth
    |-- sd-image-variations-diffusers
    |   └── unet
    |       |-- config.json
    |       └── diffusion_pytorch_model.bin
    |-- image_encoder
    |   |-- config.json
    |   └── pytorch_model.bin
    └── sd-vae-ft-mse
        |-- config.json
        └── diffusion_pytorch_model.bin
    
    

    workflow demo

    https://github.com/TMElyralab/Comfyui-MusePose/blob/main/musepose-workflow-demo.json

    https://github.com/TMElyralab/Comfyui-MusePose/assets/114042542/9cd8b9b8-6876-4281-b7a0-a7fbcb2de7e1

    Run ComfyUI workflows without the setup

    No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

    Learn more