[Please remember that is a draft under verification]
Please be sure to read this to the end for how to use this model.
This is a yolo model for the detection penis.
The detection itself is performed using a square shape, similar to how faces are detected, because this model is trained based on yorov8n.pt.

What do you use it for?
The purpose of using this model is not to swap the penis, but rather to exclude it.
When performing face swapping on an image like the one above using ADetailer, the penis gets cut off.
To resolve this issue, I enlisted the help of ChatGPT to create a mechanism that excludes the penis detected by the model from the face-swapping process.
As a result, I was able to achieve the following face-swapping effect.

While the detection accuracy itself is high, the rectangular detection area causes some issues when excluding objects. Additionally, the swapping process is not perfect, so failures happen frequently.
Please note that this version is a draft, so please bear with any issues regarding its quality.
Note that if you want to generate the image shown above using this YOLO model, modifying ADetailer is mandatory. I think ChatGPT could easily handle that for you.
I am currently at a stage prior to even considering the provision of a modified ADetailer.
[How to enable this model?]
Copy this file to <SD installation folder>\webui\models\adetailer
Using this model alone will only allow you to replace the penis.
Moreover, the outcome is unpredictable. hahaha......
[How about the modification of ADetailer?]
In my case, when I asked ChatGPT to implement a mechanism to protect the areas detected by this YOLO model, it added the following menu in "Detection".

Description
Draft, it might works or not works as expected.


