Movements Recognition in the Human Body Based on Deep Learning Strategies

Abstract

These days, the study of human body movements for the purpose of emotion identification is an absolutely necessary component of social communication. Several different contexts call for the implementation of non-verbal communication strategies such as gestures, eye movements, facial expressions, and body language. Among them, emotion detection based on body movements. It can also identify the emotions of a person even if they are too far away from the camera. Other studies have shown that body language can express emotional states more effectively than words can. In this research study, an emotional state is determined by the human motion of the entire body. The architecture of a deep convolution neural network is used, and multiple parameter settings are considered. Both the University of York's emotion dataset, which includes 15 different kinds of emotions, and dataset of GEMEP corpus, which includes five emotions, can be used to assess the proposed system. The results of the experiments demonstrated that the proposed system has a higher degree of recognition accuracy.