Aihouserender brings Stable Diffusion image generation into Blender for producing house interior renders. It works with the ADE20K dataset, a scene-parsing dataset containing annotations for objects in scenes, including house interiors. Inside Blender, the add-on loads Stable Diffusion and Lora models, preprocesses ADE20K images by segmenting them for interior design elements, and runs Stable Diffusion on those segmented images so the masks guide the generated result.
The workflow uses ADE20K as an input for Stable Diffusion's segmentation ControlNet to render a house interior. Preparing the ADE20K segmentation map, configuring ControlNet, and rendering the scene with Stable Diffusion are all implemented inside Blender. The add-on supports 3 ControlNet units, Lora, Interrogate, and ADetailer within Blender.
Aihouserender requires Stable Diffusion WebUI (AUTOMATIC1111) to be installed first, along with ControlNet in Stable Diffusion WebUI. It runs only on the Windows operating system; Linux and macOS are not supported. Documentation and tutorials are provided by the developer, and a demo video is available.