ComfyUI_SVFR

★ 99

视频修复人脸重建自动上色图像修补

ComfyUI_SVFR 是一个面向人脸视频修复的统一框架节点，支持人脸重建(BFR)、上色和图像修补等任务，便于在 ComfyUI 中一键集成并批量处理视频帧。

💡 在 ComfyUI 中批量对人脸视频进行修复、上色或补帧处理。

🍴 6 Forks💻 Python🔄 2025-03-12

🔗 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/c1eafc754fbb

📦 requirements.txt

accelerate
diffusers
moviepy
numpy
omegaconf
opencv-python
scikit-video
transformers

📄 README

ComfyUI_SVFR

SVFR is a unified framework for face video restoration that supports tasks such as BFR, Colorization, Inpainting，you can use it in ComfyUI

Update

24/02/12 修改模型加载模式为单体模型，此模型调用cofmy的vaae会偏色严重，所以只能用diffuser方法了/Change the model loading mode to a monolithic model

1. Installation

In the ./ComfyUI /custom_node directory, run the following:

git clone https://github.com/smthemex/ComfyUI_SVFR.git

2. Requirements

pip install -r requirements.txt

3. Models Required

3.1 download SVFR checkpoints from google drive After decompression, place the model in the following file format，从谷歌云盘下载模型，解压后按以下文件格式放置模型；

├── Comfyui/models/SVFR/
|   ├── id_linear.pth
|   ├── insightface_glint360k.pth
|   ├── unet.pth
|   ├── yoloface_v5m.pt

3.2 svd_xt.safetensors or svd_xt_1_1.safetensors

├── Comfyui/models/checkpoints/
|   ├── svd_xt.safetensors or svd_xt_1_1.safetensors
├── Comfyui/models/vae/
|   ├──svd.ave.fp16.safetensors  #rename from stabilityai/stable-video-diffusion-img2vid-xt/vae 重命名repo下的vae模型既可以，不命名也行

4 Inference mode

“bfr,colorization,inpainting,bfr_color,bfr_color_inpaint”,inpainting and bfr_color_inpaint mode need a mask(use comfyUI mask or black/white jpg)

5 Example

6 Citation

@misc{wang2025svfrunifiedframeworkgeneralized,
      title={SVFR: A Unified Framework for Generalized Video Face Restoration}, 
      author={Zhiyao Wang and Xu Chen and Chengming Xu and Junwei Zhu and Xiaobin Hu and Jiangning Zhang and Chengjie Wang and Yuqi Liu and Yiyi Zhou and Rongrong Ji},
      year={2025},
      eprint={2501.01235},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2501.01235}, 
}