ComfyUI_MangaNinjia

ComfyUI_MangaNinjia
★ 58

线稿上色参考图跟随漫画风着色ComfyUI节点
ComfyUI 节点,基于 MangaNinja 方法对线稿进行上色,通过精确参考图跟随实现细节一致的漫画风着色。
💡 用参考图精确给漫画线稿上色,保持配色和细节一致。
🍴 2 Forks💻 Python🔄 2025-04-09
📦
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
#gradio==3.39.0
#gradio-client==1.3.0
#h5py==3.11.0
#huggingface-hub==0.24.6
imageio
imageio-ffmpeg
importlib-metadata
importlib-resources
ipdb
ipython
ipywidgets
kornia
#kornia-rs==0.1.5
omegaconf
opencv-python
pandas
pillow
scikit-image
scikit-learn
scipy
torch
torchaudio
torchmetrics
torchvision
tqdm
transformers
einops
basicsr==1.3.5
📄 README

ComfyUI_MangaNinjia

ComfyUI_MangaNinjia is a ComfyUI node of MangaNinja which‌ is a ” Line Art Colorization with Precise Reference Following ” method。

update

  • 改进参考点的前端方法,只需要在参考和线稿图上点选对应位置的点即可,唯一要注意的是,2个节点的markimage_image栏里的canvas_image.png名称必须不一样,比如有个是canvas_image.png,另一个需要改成canvas_image1.png或则其他,请使用json工作流
  • To improve the front-end method of reference points, simply click on the corresponding points on the reference and line draft images. The only thing to note is that the name of the canvas_image.png in the markimage_image column of the two nodes must be different. For example, one node needs to be named canvas_image.png, while the other needs to be changed to canvas_image1. png or something else. Please use the JSON workflow
  • 1. Installation

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

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

    2. Requirements

  • no need, because it’s base in sd1.5 and controlnet,Perhaps someone may be missing the library.没什么特殊的库,懒得删了
  • pip install -r requirements.txt

    3. Models

  • sd1.5 address v1-5-pruned-emaonly.safetensors #example
  • controlnet lineart address control_v11p_sd15_lineart_fp16.safetensors #example
  • Annotators address sk_model.pth #example
  • MangaNinjia address #example
  • clipvison and text address #clip-vit-large-patch14
  • --  ComfyUI/models/checkpoints
        |-- any sd1.5 safetensors #任意sd1.5模型,注意要带vae的
    --  ComfyUI/models/controlnet
        |-- control_v11p_sd15_lineart_fp16.safetensors or control_v11p_sd15s2_lineart_anime_fp16.safetensors
    --  ComfyUI/models/clip
        |-- clip_l.safetensors #clip-vit-large-patch14 
    --  ComfyUI/models/MangaNinjia
            |-- denoising_unet.pth
            |-- reference_unet.pth
            |-- point_net.pthnaz
            |-- controlnet.pth
            |-- sk_model.pth

    4.Tips

  • is_lineart : if False you can link a normal image at lineart_image. 关闭is_lineart,会自动预处理图片为线稿
  • 5.Example

  • old
  • NEW
  • 6.Citation

    @article{liu2025manganinja,
      title={MangaNinja: Line Art Colorization with Precise Reference Following},
      author={Liu, Zhiheng and Cheng, Ka Leong and Chen, Xi and Xiao, Jie and Ouyang, Hao and Zhu, Kai and Liu, Yu and Shen, Yujun and Chen, Qifeng and Luo, Ping},
      journal={arXiv preprint arXiv:2501.08332},
      year={2025}
    }