Emotion perception method and system for natural human-computer interaction

Abstract

The invention discloses an emotion perception method for natural human-computer interaction. The emotion perception method comprises the following steps: S1, collecting scalp electroencephalogram signals in positive, mild and negative emotional states; s2, preprocessing the data acquired in the step S1; s3, constructing a training sample and a test sample; s4, training a deep convolutional neural network; and S5, testing the classification effect by the test set. The invention further discloses an emotion perception system for natural human-computer interaction. According to the electroencephalogram emotion recognition method, the high-dimensional spatial-temporal feature information in the electroencephalogram is learned through the spatial-temporal deep convolutional neural network, so that the efficiency and accuracy of an emotion recognition task are improved. According to the method, the problems of lost part of useful electroencephalogram feature information and feature extraction usefulness verification generated by inputting the convolutional neural network after feature extraction are solved, the complexity of emotion recognition is reduced, and the efficiency and accuracy of emotion recognition are improved.

Description

technical field [0001] The invention relates to the technical field of human-computer interaction, in particular to an emotion perception method for natural human-computer interaction. Background technique [0002] With the popularization of intelligent applications, machines play an important role in human life. The service function and sociality of machines have received extensive attention in recent years. How to improve the human-computer interaction of machines is an important topic in the field of human-computer interaction. According to emotional psychology, it is known that the user's emotional changes such as excitement, peace, and anger in the process of human-computer interaction are a kind of psychological activity characterized by the user's subjective experience. In the process of human-computer interaction, in addition to giving instructions, the user's psychological participation is more important. Through the natural human-computer interaction method based o...

AI Technical Summary

Problems solved by technology

The method used in the prior art is to extract features first and then input them into the convolutional neural network, which is prone to the problem of losing some useful EEG feature information and verifying the usefulness of feature extraction. Therefore, the deep convolutional neural network framework and its application to EEG The effectiveness of signal emotion recognition is worth exploring and researching

Images