A Yolov8 detection model that segments socks in images.
The model can be used as an ADetailer model (for Automatic1111 / Stable Diffusion use), or using other inference scripts to return detection bounding boxes of watermarks.
The model is entirely trained on the following dataset: Hello World/Computer Vision Research
A tutorial and code how to use the model can be found on this Github: https://github.com/MNeMoNiCuZ/yolov8-scripts or this CivitAI article.
Description
FAQ
Comments (4)
why is it called "Comic Speech Bubble Segmentation" when it dectects socks?
Because that's the secret identity of heroes. Socks.
It looks like I had some bad copy-pasta when I uploaded a bunch of models.
https://civitai.com/models/311872/comic-speech-bubble-detection-adetailer-model-comicspeechbubblemyolov8
There's the comic speech bubble detection model. (It's not segmentation, just detection).
This platform simply does not have enough socks. Its bad. Thank you for this. This is huge
Gotta find those smelly feet and outpaint!