Helps making the person in the right image wear the clothes in the left image.
Does not undress or make anyone nude, quite the opposite.
Add the following prompt words: Dress the figure in the right image in the clothes from the left image.
Not my lora , just re upload here with wf ,attached workflows that worked
Input text might be messed up , it does the opposite side
Description
FAQ
Comments (9)
Don't know why, it doesn't work
try "_stitch" workflow , I think it works better
Works maybe 33% of the time and you lose the likeness of the person.
For me it's doing nothing at all. it's just retaining the left image with no change at all. Same problem i have in default kontext workflows too. Kontext works great working on single images, but i never been able to get combining of two images to work. I'm using wrong prompting or kontext is just terrible at combining images. I even used the workflow provided in one of the images for this lora. and exact prompt example provided.
zengrath 我没有使用这个lora,我自己搭建了一个工作流,一个是模特去衣,一个是服装去人,就这样,让模特穿衣的提示词也只有50%以下的几率可以成功。成功率稍微高一点的是给模特画遮罩。这样它识别遮罩区域的几率更高一些。
我复制了作者提供的工作流节点，在安装好各种缺失的节点后，成功执行了换衣的操作。成功率还很高，特别是如果服装预先将人物处理掉，成功率会更高。而且在触发词后还可以继续加入调整人物动作姿态的语句，让人物动作不显得呆板
Where can I get the workflow provided by the author?
in "training data" on the right
It doesn't work most of the time, I tried many combinations.