Z-Image Turbo ControlNet
Z-Image Turbo ControlNet generates images using a written description and structural information extracted from a reference image. The reference establishes spatial or geometric guidance, while the text describes the desired appearance and content of the result.
Depth guidance captures near-and-far relationships in a scene. Edge guidance follows contours and boundaries, and pose guidance uses human body positions. The influence of this structural information is adjustable, so a reference can provide loose direction or exert stronger control over the generated composition.
What it's good for
For architectural concepts, depth guidance helps retain a scene's spatial arrangement while exploring materials or visual treatments. Edge guidance suits developing sketches into rendered illustrations and testing styles around established shapes. Pose guidance is useful for character artwork based on a figure reference. These applications give artists a concrete starting structure for exploring different appearances.
