This is an implementation of a background removal models trained using the Highly Accurate Dichotomous Image Segmentation research for use with InvokeAI 4.0 and above.
Supported Models:
- BriaAI - A SOTA background removal model from BriaAI.
- Open RMBG - A fully open-source model optimized for images with humans. This model is similar to BriaAI 1.4, but with open training data/process and commercially free to use.
Note: If you are looking to use this with an older version of Invoke, please use the 3.x branch instead.
Go to your nodes folder in your root directory and clone this repo
git clone https://github.com/blessedcoolant/invoke_bria_rmbg.git
The required models will be automatically downloaded by Invoke's Model Download API.
Add the BRIA AI Background Removal node. Plug in an image. That's it.
Feel free to fix anything or improve it anyway and make a PR. Thank you!