Development of the software to improve the accuracy of emotion recognition Available to Purchase

The theory of pattern recognition is one of the directions of scientific and technical development in the field of image analysis and machine vision, etc. The authors consider the possibility of using information systems to improve the quality and accuracy of recognizing emotions on faces from images. The purpose of the research is to develop an application that uses the most effective algorithm for detecting faces and recognizing emotions. The methodology is based on the principles of deep learning, includes the use of convolutional neural networks, their software embodiments. The functional scheme is implemented in the application using the methods of the library of computer vision. Result of the research, a module was developed with the ability to determine the user’s emotions from an image from a webcam in real time. At the testing stage a data set for training was formed. A convolutional neural network was trained. The functional for generating graphs describing the emotional portrait of the user for the entire time of using the module was implemented. In the future, it is planned to improve the quality of neural network recognition, and prepare a web service for integration into mobile applications.