A technique that divides an image into small rectangular sections (patches) and processes each one separately, rather than treating the entire image as a single unit.
Quality of vision, audio, and image understanding (distinct from modality support)