: New optical-flow-based tools allow for "deep" image patching and cleanup, such as applying digital makeup that perfectly follows the contours and depth of a face.

: "Deep" tools are GPU-intensive. For optimal performance, a system with Apple Silicon or a dedicated GPU with at least 4GB to 8GB of VRAM is recommended.

: The DaVinci Neural Engine can automatically generate a 3D depth map of a 2D scene. This enables you to:

DaVinci Resolve Studio 20 and 21 have introduced capabilities, specifically focusing on advanced deep image compositing and AI-driven spatial analysis . These tools allow for complex visual effects work where multiple depth layers are stored per pixel, moving beyond simple 2D depth maps to true 3D spatial compositing. Key "Deep" Post-Production Features