ComfyUI-PhotoMaker-Plus

ComfyUI-PhotoMaker-Plus
★ 296

图像生成人脸重现可定制触发词ComfyUI扩展
为ComfyUI提供PhotoMaker/PhotoMaker-V2模型参考实现，支持onnxruntime与insightface，提升人脸相似度并可自定义多触发词，附带样式与预处理节点。
💡 在ComfyUI中使用PhotoMaker（含V2）生成高相似度人脸并自定义触发词。
🍴 23 Forks💻 Python🔄 2024-12-01
📦
网盘下载
复制链接后前往夸克网盘下载
https://pan.quark.cn/s/86a6deb1b5f6
📦 requirements.txt
onnxruntime
onnxruntime-gpu;
sys_platform
!=
'darwin'
and
platform_machine
==
'x86_64'
insightface
numpy<2
📄 README

ComfyUI PhotoMaker Plus

ComfyUI implementation for PhotoMaker.

PhotoMaker implementation that follows the ComfyUI way of doing things. The code is memory efficient, fast, and shouldn’t break with Comfy updates.

Installation

  • Install ComfyUI.
  • Install onnxruntime and insightface.
  • Install through ComfyUI Manager or clone this repo into custom_nodes by running the following commands in a terminal:

    • “`bash

    cd ComfyUI/custom_nodes

    git clone https://github.com/shiimizu/ComfyUI-PhotoMaker-Plus.git

    “`

  • Download the model(s) from Hugging Face (V1, V2) and place it in a photomaker folder in your models folder such as ComfyUI/models/photomaker.
  • Check out the example workflows.

    • Features of this Plus version

  • Better face resemblance by using CLIPImageProcessor like in the original code.
  • Customizable trigger word
  • Allows multiple trigger words in the prompt
  • Extra nodes such as PhotoMakerStyles and PrepImagesForClipVisionFromPath

    • Important news

    2024-09-01

  • A PhotoMakerLoraLoaderPlus node was added. Use that to load the LoRA.

    • 2024-07-26

  • Support for PhotoMaker V2. This uses InsightFace, so make sure to use the new PhotoMakerLoaderPlus and PhotoMakerInsightFaceLoader nodes.

    • 2024-01-24

  • Official support for PhotoMaker landed in ComfyUI. Therefore, this repo’s name has been changed. The PhotoMakerEncode node is also now PhotoMakerEncodePlus.

    • 2024-01-21

  • Due to various node updates and changes, you may have to recreate the node.
  • Removed ref_images_path input. Added a PrepImagesForClipVisionFromPath node.

    • 2024-01-18

  • No need to manually extract the LoRA that’s inside the model anymore.

    • Citation

    @article{li2023photomaker,
  title={PhotoMaker: Customizing Realistic Human Photos via Stacked ID Embedding},
  author={Li, Zhen and Cao, Mingdeng and Wang, Xintao and Qi, Zhongang and Cheng, Ming-Ming and Shan, Ying},
  booktitle={arXiv preprint arxiv:2312.04461},
  year={2023}
}