adetailer/README.md

!After Detailer

!After Detailer is a extension for stable diffusion webui, similar to Detection Detailer, except it uses ultralytics instead of the mmdet.

Model

Model	Target	mAP 50	mAP 50-95
face_yolov8n.pt	2D / realistic face	0.660	0.366
face_yolov8s.pt	2D / realistic face	0.713	0.404
mediapipe_face_full	realistic	-	-
mediapipe_face_short	realistic	-	-

The yolo models can be found on huggingface Bingsu/adetailer.

Dataset

Datasets used for training the yolo face detection models are:

• roboflow AN
• roboflow Anime Face CreateML
• roboflow xml2txt
• wider face

User Model

Put your ultralytics model in webui/models/adetailer. The model name should end with .pt or .pth.

It must be a bbox detection model and use only label 0.

ControlNet Inpainting

You can use the ControlNet inpaint extension if you have ControlNet installed and a ControlNet inpaint model.

On the ControlNet tab, select a ControlNet inpaint model and set the model weights.

Example

Changelog

2023-05-02

• Remove None from model list and add Enable ADetailer checkbox.
• install.py skip_install fix.