Analyzing Infant Crying Patterns: Classification of Hunger and Discomfort
No Thumbnail Available
Date
2021
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Uva Wellassa University of Sri Lanka
Abstract
Infant crying is the crying of infants as a response to an internal or external stimulus. Infants cry as a
form of basic instinctive communication. An infant's cry contains a lot of information about the baby
such as hunger, pain, discomfort, sleepiness, burp, anger, etc. Parents‟ or guardian's inability to
recognize and timely address the reason for the infant's cry prompts dissatisfaction for the infant and
a feeling of helplessness for the parents. Therefore, an accurate, efficient automatic method for
analyzing infant cry patterns and notifying the cause of cry is a very useful tool for parents. This
study aims at the detection of baby cry patterns and identification of uniqueness of the hunger and
discomfort crying patterns of the infants. This is achieved through analyzing the different patterns of
the sound waves of the infants crying by converting the crying signal to an equivalent frequency
waveform. This novel model can fill the gaps in the current models by achieving higher accuracy.
Thus, this research is beneficial in various ways such as reducing parental dissatisfaction and
helplessness when infant crying, minimizing child abuse and helping parents to better understand
their child‟s needs and psychosis by analyzing crying patterns. The proposed model detects sound
frequency, draws the waveform of the signal, and uses a Convolutional Neural Networks
methodology to identify and distinguish the crying patterns of the infants. The dataset was collected,
and the crying patterns were labeled by getting assistance from the domain experts. The model was
trained and tested against the labeled data and it gained an accuracy of 91%. The proposed model will
be further enhanced to identify more feelings of the infants and will be further developed to assists to
recognize selected medical conditions by analyzing the crying patterns of the infants.
Keywords: Cry signal; Discomfort; Hunger; Infant crying patterns
Description
Keywords
Social Science, Computing and Information Science, Communication Technology, Health Science