In this video, I am showing a demo usage of a Python script that connects Blender into WebUI Stable Diffusion running ControlNet extension, allowing you to render an image, and process it with "depth" data and generate a new imagery based on your prompt!
The Script:
https://github.com/coolzilj/Blender-C...
As for this demonstration, I am using an AI model called ZUKKI that has been trained using "Ma vie de courgette" stop motion animation animation which perfectly works with our Nelb Rig:
https://huggingface.co/tuwonga/zukki_...
For future application, I believe the AI with ControlNet could help in rendering / improving a basic rig and adding "details" based on the prompt and trained model.
Requirements:
Good Machine that can run AI stable diffusion. I am using M2 Pro Machine.
Stable Diffusion WebUI by Automatic 1111
Nelb Rig:
https://cgcookie.com/downloads/red-ne...