IPAdapter Regional Conditioning

利用IPAdapter Regional Conditioning的风格迁移和通过遮罩构图，让AI生成的图片更加具有可控性，缺点是要得到构图和细节都完美的图片目前还是需要非常多的尝试，

Using the style transfer of IPAdapter Regional Conditioning and the mask composition, the pictures generated by AI are more controllable. The disadvantage is that it still needs a lot of attempts to get perfect pictures with perfect composition and details.

V1.Regional Conditioning 3

可以将2个主体与一个背景进行结合，得到一张重新构图的图片，

You can combine two subjects with a background to get a recomposed picture,

整个工作流加载的模型比较多，对低配电脑不太友好，熟练使用comfyui的可以把提示词提取和面部细化部分单独放到新的工作流中使用，以减少显存占用，

The whole workflow loads a lot of models, which is not friendly to low-power computers. Those who are skilled in using comfyui can use the hint word extraction and face refinement separately in the new workflow to reduce the GPU memory occupation

V2.Regional Conditioning 2

将1个主体与一个背景进行结合，得到一张重新构图的图片。类似局部重绘，但是通过IPAdapter引入了新的风格，让图片呈现出一些不一样的东西。缺点是IPAdapter目前的通病，就是会让新图片的细节出现缺失。但如果用来构图的话，则还是相当有价值的。

Combine a subject with a background to get a recomposed picture. Similar to partial redrawing, but with the introduction of a new style through the IPAdapter, making the image appear something different. The disadvantage is that IPAdapter is currently a common problem, that is, the details of the new picture are missing. But if it is used for composition, it is still quite valuable.