Machine Perception (La Perception de l'Environment par Senseurs Automatiques).

Human perceptual capabilities involve the extraction of task-oriented information from environmental stimuli through physical sensing and the use of background knowledge. There are many activities underway aimed at providing similar capabilities of artificial machine perception. Some success is achieved by exploiting what is known of corresponding human cognitive processes and by making use of the increasing power of information processing techniques. For this purpose, the recognition of sharply contrasted as well as fuzzy patterns stationary or dynamically changing plays an important role along with other aspects of processing of complex information structures. These techniques are beginning to be applied in guidance and control, in particular with regard to artificial visual perception and speech understanding. This application promises major benefits with the advent of autonomous vehicle and mission control, and of intelligent systems for situation awareness support of human operators. This Lecture Series covers the following subjects Pattern recognition techniques Real time visual machine perception, principles and applications in G and C Real time speech recognition and understanding in the G and C domain. This Lecture Series, sponsored by the Guidance and Control Panel of AGARD, has been implemented by the Consultant and Exchange Programme.