Deep Learning Assisted Neonatal Cry Classification via Support Vector Machine Models.

Ashwini K,Kathiravan Srinivasan,Chuan-Yu Chang,P. M. Durai Raj Vincent

doi:10.3389/fpubh.2021.670352

Abstract

Neonatal infants communicate with us through cries. The infant cry signals have distinct patterns depending on the purpose of the cries. Preprocessing, feature extraction, and feature selection need expert attention and take much effort in audio signals in recent days. In deep learning techniques, it automatically extracts and selects the most important features. For this, it requires an enormous amount of data for effective classification. This work mainly discriminates the neonatal cries into pain, hunger, and sleepiness. The neonatal cry auditory signals are transformed into a spectrogram image by utilizing the short-time Fourier transform (STFT) technique. The deep convolutional neural network (DCNN) technique takes the spectrogram images for input. The features are obtained from the convolutional neural network and are passed to the support vector machine (SVM) classifier. Machine learning technique classifies neonatal cries. This work combines the advantages of machine learning and deep learning techniques to get the best results even with a moderate number of data samples. The experimental result shows that CNN-based feature extraction and SVM classifier provides promising results. While comparing the SVM-based kernel techniques, namely radial basis function (RBF), linear and polynomial, it is found that SVM-RBF provides the highest accuracy of kernel-based infant cry classification system provides 88.89% accuracy.

Highlights

The neural network has an input, two hidden, and an output layer to calculate the weighted prosodic features. Those features were taken as input to the deep learning approach, which is found that the merged features can distinguish the variation present in infant cry signals
The audio cry signals are converted into a spectrogram image using short-time Fourier transform (STFT)
The spectrogram images are fed into the deep convolutional network

Summary

INTRODUCTION

Jam and Sadjedi [4] carried out work to distinguish the pain and normal infant cries They observed that, while processing the audio signal, silence elimination, filtering, pre-emphasizing was crucial. Support vector machine (SVM) and expectation-maximization (EM) algorithms over an expert system were employed to classify the data It shows that nonlinear feature with an expert system-based classification approach gives better performance [5]. Implementing the deep learning approach needs millions of data samples to get the best results This motivates us to enhance the infant cry classification model’s performance even with the small dataset by extracting the features using the deep learning technique and classifying the infant cries using a machine learning algorithm with less computational complexity. This work classifies the most common infant cries such as hunger, pain, and sleepiness

RELATED WORKS

RESULTS AND DISCUSSION

CONCLUSION

ETHICS STATEMENT