YES NO

Computer vision provides nowadays powerful tools for the interaction between man and computer. Low cost video image processing systems make it possible for a human to use head, hands, fingers as digital input devices, without touching a keyboard, a mouse or other commonly used peripheral devices. In such an augmented reality, information is acquired by observing the way subject behaves. A simple example of an augmented reality is provided by the presented system which recognizes "yes " or “no ” through the movement of a subject’s head. In this abstract an experimental setup enabling a simple communication between the human and computer system is presented. Vision of the face is an important aspect of human-to-human communication. We have been experimenting with the use of computer vision to "watch the face". In this contribution we propose a robust technique for detecting, tracking and recognizing images of face movements. When combined with real time image processing and active control of camera parameters, these technique can greatly improve the communications between human and computer in special applications, in which the interaction using hands, fingers or voice is not desired or impossible. The system is depicted in Figs. 1 and 2, in which the face detection and object tracking modules i