In this project i tried to combine AI-generated Images from stable diffusion with 3d. I made a reference image. Then i modeled basic shapes in a 3d blockout. Back in stable diffuion, i genereted textures(grass, stoneplates, bricks), and hole objects(trees, plants, windows)
It was a lot of testing different promts to get good results, in a repetitive style. Some images are cutout by masks or by hand using the knife tool.
This was a tes to see how good i can combine these tools, and recently i tested different ideas in comic style, so to have scenes and background in 3d is possible, mut maybe lowpoly modeling with simple shaders is easier and faster???
Also tested some outlines with freestyle and solidify modifiers, but was not great in this example.
Small animation on Artstation: https://www.artstation.com/artwork/qez3Dz
Oh wow that looks so pretty, it's like The Spice Vendor's summer vacation house. I love how lovely it looks, the colors and the lighting. Kudos for going all in on embracing A.I., that's the future. Sarah Connor is looking at you though. Awesome project Tobias 🤙
the matrix is also waiting to open its rabbit hole =)