%0 Journal Article
%T Multilevel Cognitive Machine-Learning-Based Concept for Artificial Awareness: Application to Humanoid Robot Awareness Using Visual Saliency
%A Kurosh Madani
%A Dominik M. Ramik
%A Cristophe Sabourin
%J Applied Computational Intelligence and Soft Computing
%D 2012
%I Hindawi Publishing Corporation
%R 10.1155/2012/354785
%X As part of ¡°intelligence,¡± the ¡°awareness¡± is the state or ability to perceive, feel, or be mindful of events, objects, or sensory patterns: in other words, to be conscious of the surrounding environment and its interactions. Inspired by early-ages human skills developments and especially by early-ages awareness maturation, the present paper accosts the robots intelligence from a different slant directing the attention to combining both ¡°cognitive¡± and ¡°perceptual¡± abilities. Within such a slant, the machine (robot) shrewdness is constructed on the basis of a multilevel cognitive concept attempting to handle complex artificial behaviors. The intended complex behavior is the autonomous discovering of objects by robot exploring an unknown environment: in other words, proffering the robot autonomy and awareness in and about unknown backdrop. 1. Introduction and Problem Stating The term ¡°cognition¡± refers to the ability for the processing of information applying knowledge. If the word ¡°cognition¡± has been and continues to be used within quite a large number of different contexts, in the field of computer science, it often intends artificial intellectual activities and processes relating ¡°machine learning¡± and accomplishment of knowledge-based ¡°intelligent¡± artificial functions. However, the cognitive process of ¡°knowledge construction¡± (and in more general way ¡°intelligence¡±) requires ¡°awareness¡± about the surrounding environment and, thus, the ability to perceive information from it in order to interact with the surrounding milieu. So, if ¡°cognition¡± and ¡°perception¡± remain inseparable ingredients toward machines intelligence and thus toward machines (robots¡¯, etc.) autonomy, the ¡°awareness¡± skill is a key spot in reaching the above-mentioned autonomy. Concerning most of the works relating modern robotics, and especially humanoid robots, it is pertinent to note that they either have concerned the design of controllers controlling different devices of such machines [1, 2] or have focused the navigation aspects of such robots [3¨C5]. In the same way, the major part of the work dealing with human-like, or in more general terms intelligent, behavior, has connected abstract tasks, as those relating reasoning inference, interactive deduction mechanisms, and so forth. [6¨C10]. Inspired by early-ages human skills developments [11¨C15] and especially human early-ages walking [16¨C19], the present work accosts the robots intelligence from a different slant directing the attention to emergence of ¡°machine awareness¡± from both ¡°cognitive¡± and ¡°perceptual¡± traits. It is
%U http://www.hindawi.com/journals/acisc/2012/354785/