Infant Cry Analysis for Emotion Detection by using Cross-Correlation Algorithm

Abstract

Crying is the part of normal behavior of human. It is the first way of communication for the infant after his birth and it is a very positive sign of a new healthy life. There are several reasons for which infant cry including hunger, discomfort, and need for attention. Analysis of infant cry is very useful because infant cry contains useful information, which helps in detecting the nature of infant cry and the reason due to which he/she is crying. Information inside the crying signal is used to detect the infant’s psychological condition. The analysis of the infant cry easily deduces the state of health of newborn babies. Many research papers have been written on the analysis of infants by using various methods i.e., spectrography, melody shape method, inverse filtering, etc. In this paper we used a different method to detect the emotion of crying infant. Cross-correlation algorithm is used to detect the reason due to which infant is crying. GUI (Graphical User Interface) is developed to study and analyze the reason. In the present work the study is performed for five a reason due to which infant is crying: hascolic, hungry, sad, stomach pain, unhappy.

Keywords: Infant cry, GUI, Cross-correlation, STE (short term energy)