ComfyUI_RH_DreamID-V/README.md

ComfyUI_RH_DreamID-V

A ComfyUI plugin for DreamID-V, enabling high-fidelity video face swapping powered by Diffusion Transformer technology.

✨ Features

• 🎭 High-Fidelity Face Swapping: Advanced video face swapping using Diffusion Transformer
• 🎬 Video-Driven: Use video as motion/pose driver
• 🖼️ Reference Image: Single face image as identity reference
• 🔧 ComfyUI Integration: Seamlessly integrated into ComfyUI workflows

📋 Nodes

This plugin provides two core nodes:

Node Name	Description
RunningHub_DreamID-V_Loader	Load the DreamID-V model pipeline
RunningHub_DreamID-V_Sampler	Execute video face swapping sampling

🛠️ Installation

Method 1: Via ComfyUI Manager (Recommended)

1. Install ComfyUI Manager
2. Search for ComfyUI_RH_DreamID-V in ComfyUI Manager
3. Click Install

Method 2: Manual Installation

1. Navigate to ComfyUI's custom_nodes directory:

cd ComfyUI/custom_nodes

2. Clone this repository:

git clone https://github.com/HM-RunningHub/ComfyUI_RH_DreamID-V.git

3. Install dependencies:

cd ComfyUI_RH_DreamID-V
pip install -r requirements.txt

📦 Model Downloads & Configuration

This plugin requires the following model files (refer to Official Model Preparation Guide):

Models	Download Link	Notes
DreamID-V	🤗 Huggingface	Supports 480P & 720P
Wan-2.1	🤗 Huggingface	VAE & Text encoder

1. Wan2.1-T2V-1.3B Base Model

Download from: 🤗 Huggingface - Wan2.1-T2V-1.3B

Place the files in the following directory:

ComfyUI/models/Wan/Wan2.1-T2V-1.3B/
├── models_t5_umt5-xxl-enc-bf16.pth
├── Wan2.1_VAE.pth
└── google/umt5-xxl/  (tokenizer folder)

2. DreamID-V Model

Download from: 🤗 Huggingface - DreamID-V

Place the file in the following directory:

ComfyUI/models/DreamID-V/
└── dreamidv.pth

Complete Model Directory Structure

ComfyUI/
└── models/
    ├── Wan/
    │   └── Wan2.1-T2V-1.3B/
    │       ├── models_t5_umt5-xxl-enc-bf16.pth
    │       ├── Wan2.1_VAE.pth
    │       └── google/
    │           └── umt5-xxl/
    └── DreamID-V/
        └── dreamidv.pth

🚀 Usage

1. Add the RunningHub_DreamID-V_Loader node in ComfyUI to load the model
2. Add the RunningHub_DreamID-V_Sampler node
3. Connect the following inputs:
   • pipeline: Model pipeline from the Loader node
   • video: Driving video (containing motion/pose)
   • ref_image: Reference face image
4. Configure parameters:
   • size: Output size (832480 or 1280720)
   • frame_num: Number of frames (must be 4n+1, e.g., 81)
   • sample_steps: Sampling steps (default: 20)
   • fps: Frame rate (default: 24)
   • seed: Random seed

⚙️ Parameters

Parameter	Description	Default
size	Output video resolution	832*480
frame_num	Number of output frames (4n+1)	81
sample_steps	Diffusion sampling steps	20
fps	Output video frame rate	24
seed	Random seed	42

💻 System Requirements

• GPU: NVIDIA GPU with VRAM >= 16GB recommended
• Python: 3.8 or higher
• CUDA: 11.7 or higher
• ComfyUI: Latest version

📝 Dependencies

• torch >= 2.0.0
• torchvision >= 0.15.0
• easydict
• numpy
• Pillow
• opencv-python
• decord
• tqdm
• mediapipe

🙏 Acknowledgements

• DreamID-V - Original project by ByteDance
• Wan Team - Wan video generation model
• ComfyUI - Powerful Stable Diffusion GUI

📄 License

This project is licensed under the Apache-2.0 License.

⚠️ Disclaimer

This project is for educational and research purposes only. Please ensure compliance with relevant laws and regulations when using this tool. Do not use it for illegal purposes or to infringe upon the rights of others.

---

If you find this project helpful, please give it a ⭐ Star!