Conditioning Stable Diffusion with Segment Anything
Please view our blog post for details about our project
Notebooks contains the 3 main notebook used during our project.
The notebook should be run in that order
- Sam-Controlnet-Dataset
- Sam-Controlnet-Train
- Sam-Controlnet-Eval
This create a custom controlnet trained on the segment anything model It is then possible to use that controlnet model to condition stable diffusion with segmentation masks