This tool helps you train detection models, as well as use them to generate detection outputs (image and text).

It can be used inside Automatic1111 or ComfyUI with the right extensions, like ADetailer or similar node-packs.

The file can be downloaded from this page, or cloned from Github:

https://github.com/MNeMoNiCuZ/yolov8-scripts/

I have also written an article to help you create your first model:

https://civarchive.com/articles/4080