欢迎进我的comfyui交流QQ群：971213362

推特(X)：https://x.com/AIdugumatai

youtube：https://www.youtube.com/channel/UCfQnrAw5yQQmBU04I4aqh9A

Workflow steps:

1. Load clothing images and character images

2. Load large models, use Gemini 1.5 pro to identify clothing images and infer simple prompt words

3. Process masks throughout the entire workflow

4. Generate works

5. Second production of works

6. Slightly repair clothing

7. Losslessly enlarge the final image

The workflow needs to load two checkpoints, a regular SDXL model and an inpainting version of the SDXL model

The first one: Juggernaut XL inpainting

https://civitai.com/models/403361?modelVersionId=449759

The second one: any SDXL large model

The main plug-ins used are:

ComfyUI-Gemini

LayerStyle

comfyui-mixlab-nodes

ComfyUI Easy Use

BrushNet

Other model downloads:

BrushNet: https://drive.google.com/drive/folders/1fqmS1CEOvXCxNWFrsSYd_jHYXxrydh1n

Put it in ComfyUI\models\inpaint\BrushNet

segformer_b2_clothes: https://huggingface.co/mattmdjaga/segformer_b2_clothes

Put it in ComfyUI/models/segformer_b2_clothes

↓

For more details about this workflow, please see the comments in the workflow!

工作流程步骤：

1.加载服装图像和人物图像

2.加载大模型，利用Gemini 1.5 pro 识别服装图像推理出简单的提示词

3.处理贯穿整个工作流程的遮罩

4.生成作品

5.作品第二次制作

6.对服装进行略微的修复

7.对最终的图像进行无损放大

工作流需要加载两个大模型，一个常规的SDXL模型，一个inpainting版的SDXL模型

第一个：Juggernaut XL inpainting

https://civitai.com/models/403361?modelVersionId=449759

第二个：任何SDXL大模型均可

主要用到的插件有：

ComfyUI-Gemini

LayerStyle

comfyui-mixlab-nodes

ComfyUI Easy Use

BrushNet

其他模型下载：

BrushNet：https://drive.google.com/drive/folders/1fqmS1CEOvXCxNWFrsSYd_jHYXxrydh1n

将其放入 ComfyUI\models\inpaint\BrushNet

segformer_b2_clothes：https://huggingface.co/mattmdjaga/segformer_b2_clothes

将其放入 ComfyUI/models/segformer_b2_clothes

↓

关于此工作流的更多介绍详情请看工作流中的注释释