Music Therapy-Based Emotion Regulation Using Convolutional Neural Network

Music Therapy-Based Emotion Regulation Using Convolutional Neural Network

Ladly Patel
Copyright: © 2022 |Pages: 24
DOI: 10.4018/978-1-7998-7776-9.ch003
OnDemand:
(Individual Chapters)
Available
$33.75
List Price: $37.50
Current Special Offers
10% Discount:-$3.75
TOTAL SAVINGS: $3.75

Abstract

Humans communicate with each other and express emotions using facial expressions. Facial expression is an important part of expressing emotions. Facial features can be considered as eyes, mouth, and nose. In this chapter, the authors considered these facial features for emotion detection and processed them with convolutional neural network (CNN). There are mainly six basic types of emotions: fear, disgust, anger, sadness, happiness, and surprise. These emotions can be classified into two types: positive emotions and negative emotions. A positive emotion is a feeling where there is no negativity such as happy, neutral. A negative emotion is a feeling of depression, frustration including anger, sadness, fear. This chapter describes a step-by-step method in processing an image in CNN and giving an output. CNN classifies different emotions. Further classification is done for emotion as negative and positive, and when the negative emotion is detected, music is played to change the emotion of a person.
Chapter Preview
Top

Background

The term Convolutional neural networks was first introduced in the year 1980 by Yann LeCun. Convolutional neural network (CNN) is a class of deep learning neural network. Convolutional networks was inspired by observing the patterns between neurons in animal visual cortex. A convolutional neural network is a kind of artificial neural network which is used in image recognition and then processing which is designed to process a pixel data. CNN consists of an input layer, hidden layer and an output layer. CNN extract features from the images. The input layer in CNN is a grayscale image. The output layer consists of multi-class labels. Hidden layer consists of convolution layer, ReLu layer, Pooling layer and a fully connected neural network. To train CNN, each input image is need to pass through multiple series of convolution layers with kernals, ReLu layer, Pooling layers and at last to fully connected layer so that an object or image can be classified with probability values ranging between 0 and 1. For this chapter, some survey were done to proof that CNN is best for processing and classifying a images. Some of the surveys are listed below:

Complete Chapter List

Search this Book:
Reset