generative illustrations with controlNET & img2img

With the help of customly illustrated guide maps and dedicated LORA’s, we can guide the generative result into a specific direction. This works best with a combination of basic img2img offset and a combination of suitable controlNET modules like depth or segmentation. In the following, check out some examples:

handpainted img2img base, depthmap for controlNET and prompted result

monochrome ink pen bright contrast of ( surreal detailed crochet marcrame knittings traditional weaving kinitting patterns out of cords and shaman costumes, automation artificial neuronal networks curly curvy :1.2), light grey pattern background voronoi nodes, (bright lights high key lighting:1.2) , ( sharp, sharp edges, in focus,vector:1.5 )
3d render blender drawing painting, ( female, breast, sexy, hero, face:2:0), DOF, blur, blurry, grayscale