comfyui-self-guidance

comfyui-self-guidance
★ 10

扩散模型提示增强采样控制ComfyUI 插件
ComfyUI 的非官方 Self-Guidance 实现，通过在采样过程中生成内部引导信号，提高生成图像对提示的忠实度与细节控制。
💡 在ComfyUI流程中加入自我引导以提升提示一致性与细节表现
🍴 4 Forks💻 Python🔄 2025-04-17
📦
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
diffusers==0.26.3
fuzzywuzzy==0.18.0
carrot0.8
carrot1.3
eclair0.8
eclair1.3
📄 README

comfyui-self-guidance

Unofficial ComfyUI implementation of Diffusion Self-Guidance for Controllable Image Generation

Installation

Clone this repo into custom_nodes folder or use Manager.

Workflow

the spatial manipulation of objects in real images

  • example workflow

    • Example 1

    use *example1.jpg*

  • “an onion and a carrot“
  • set “carrot` as target object, set `onion“ as fix object
  • shrink, scale factor = 0.8
  • enlarge, scale factor = 1.3

    • Example 2

    use *example2.jpg*

  • “a photo of an eclair and a shot of coffee“
  • set “eclair` as target object, set `coffee“ as fix object
  • shrink, scale factor = 0.8
  • enlarge, scale factor = 1.3

    • Note

    Currently only SD series models are supported.