Thesis (Ph.D.) - Cairo University - Faculty of Computers and Artificial Intelligence - Department of Information Technology

Humans can internally sense their muscles.They can also recognize their environment with its objects and are able to navigate through it easily to reach them. How these abilities are gained and interact each other is still an open question in biology. Our aim is to help biologists to understand how a human is able to build his cognitive map and make goal-directed movements.The origin of human capabilities goes back to the fetus stage. We present a theoretical model of the development of the cognitive map of a fetus human from his sensorimotor system.The model integrates the proprioceptions of body limbs and perceptions from the environment and how these cooperate to build a cognitive map, which in turn, is essential for making goal-directed movements to reach di erent objects in the surrounding environment.We propose a new clustering algorithm called Frequency-based-means ; which is used to get the proprioceptions and the perceptions that form the association map. Hidden Markov Models are used to model movement learning and production

Issued also as CD