thedeoxen/MOSS-Image-pose-controlnet
Image-to-Image • Updated • 5
Control LoRAs for MOSS-Image-SFT. Pose, depth, or a drawing holds the picture. The prompt does the rest.
Note Pose. The skeleton keeps the body.
Note Depth. Near and far stay put.
Note Canny. The outline stays.
Note Lineart. A drawing becomes a picture.
Note Normals. The volume stays.
Note Scribble. A rough sketch is enough.
Note Segmentation. Regions hold the layout.