The elements of a face can be hard for computers — and for some people — to recognize.

Henry Schneiderman, a computer vision researcher at Carnegie Mellon University’s Robotics Institute, has developed the most accurate program in existence for detecting faces in still images and video.

Schneiderman’s face detector uses low-resolution black and white images measuring 24 by 32 pixels. Part of the development process involves showing the computer examples of faces and non-faces and letting the computer program gradually develop its own statistical rules for determining what constitutes a face.

Key facial parameters: symmetry across the face (particularly the spacing of the eyes), the forehead, and configuration of eyes, nose and mouth.

Uses include security and a system for automatically reducing “red eye” that can show up in some flash photos.