Perspective Chapter: Emotion Detection Using Speech Analysis and Deep Learning

Abstract

Speech reflects the sentiment and emotions of humans. People can identify the emotional states in speech utterances, but there is a higher chance of perception error, which is generally termed as human error to identify the proper emotion when only using speech signals. Thus, artificial intelligence plays an important role in the detection of emotion through speech. Deep Learning is the subset of Machine Learning (ML) and artificial intelligence through which speech signal processing can be performed and the detection of emotions can be accomplished using speech. In this chapter, the classifiers of Machine Learning and Deep Learning will be reviewed. From the comparison in various studies and performances we will conclude what methods work better than others. We will discuss the limitations of these approaches as well. Accuracy scores will be discussed for each proposed system.