Methods and Models for Identifying Human Emotions by Recognition Gestures and Motion

Abstract

The paper describes a methodology, models and automated system for anidentification of human emotions based on analysis of body movements, arecognition of characteristic gestures and poses. In the model of personemotions, the typical body movements are formalized with linguisticvariables and fuzzy hypergraphs for temporal events. Emotional states ofthe real person recognized with the automated system are presented in alimited natural language. Shown the allocation of granules and relatedpostures. Compliance granules poses and basic emotional states by K.Izard is presented. The article shows how to apply system forrecognition and translation in real time gestures of the Russian deaflanguage in the text and the text in gestures. The system is intendedfor training of children with hearing disabilities and adults who needto learn sign language.