The development of an algorithmic model for object recognition from visual and sound information — Based on neuro-fuzzy logic

Abstract

This paper considers the problem of recognizing the visual and sound information by constructing a virtual environment, which allows to qualitatively simplify the system and to carry out of experiments, and to create an algorithmic model of pattern recognition comparable to human capabilities. Our research is aimed at obtaining an algorithmic model that can extract from the surrounding world “meaningful” (visual and sound) objects to link with the relevant lexical concepts, concepts which are atomic building blocks of intelligence. Our general objective is to experimentally analyze the problem of artificial intelligence in order to further the development of machine intelligence — by achieving a phase-transition-type drastic increase in the complexity of the behavior of artificial personality (AP).