Physical Activity Set Selection for Emotional State Harmonization Based on Facial Micro-Expression Analysis

A new approach to harmonize a human emotional state is proposed. It is based on timely revealing and working through hidden negative emotions. The main ideas proving the possibility to implement this approach are described. The experimental investigation results for selected methods of revealing hidden negative emotions and working them through are represented. The video sequences reproducing emotions are used as input data. The ultimate research aim is the software system construction for the human emotional state harmonization. The effect of using such a system will be maintaining the health and increasing the quality of life of the modern human. The basic functions of this system are highlighted. The problem of repressing emotions and its possible negative consequences for the health are considered. The possibility of revealing repressed emotions from facial micro-expressions is justified. The main stages of the micro-facial movement detection by hybrid methods are considered in details. Each stage results are illustrated using the software pipeline developed in the previous research. It is proposed to work through revealed negative emotions by executing the individually selected physical activity set. The problem of the execution accuracy control of these exercises when working without assistance is considered. The health-improving exercise model is represented. The possibility of recording the motion trajectory using a smartphone software solution is shown. The method of the motion execution accuracy estimation on the basis of dynamic time warping is described. The development and implementation prospects of the proposed approach based on a smartphone software solution are justified.