Emotion recognition method and system based on deep learning model and long-short memory network

Abstract

The invention discloses an emotion recognition method and system based on a deep learning generation model and a long-short memory network. The method includes performing data preprocessing on EEG signals and dividing data sets, and constructing a network model. The network model includes a variational encoder. The image reconstruction model formed and the emotion recognition model formed by the long-short memory network; the objective function is constructed according to the network model; the network model is trained by using the training set, and the objective function is optimized by using the Adam optimization operator in the neural network to obtain the trained Network model; use the cross-validation set to cross-check the trained network model, determine the hyperparameters of the network model, and obtain the final network model; use the final network model to visually generate seed data and recognize emotions. The present invention relies on the data artificial intelligence method to deeply learn the complex structure of the collected EEG signal in space and time, reduces the subjective factors in the prediction, and improves the accuracy of the prediction.

Description

technical field [0001] The invention relates to the technical field of machine learning, in particular to an emotion recognition method and system based on a deep learning model and a long-short memory network. Background technique [0002] The study of the human brain has been extensively improved in recent decades as the practical needs of normal human life have evolved. With the rapid development of human-computer interaction (HCI) and brain-computer interface (BCI), the visualization and generation of EEG signals has become particularly important through the wide application prospects of EEG signals. [0003] Emotions generated by the human brain are an important mechanism to ensure people's daily survival and adapt to the environment, and further affect people's daily study, life, work, and decision-making. However, the external observation of the human body will be affected by people's daily culture, and some scales and emotion detection will be affected by subjective...

AI Technical Summary

Problems solved by technology

[0004] Existing emotion recognition methods often adopt traditional machine learning or signal processing methods, which make it difficult to learn EEG signals with complex neural structures to identify corresponding emotions

At the same time, traditional methods do not have the ability to visualize and generate signals

Images