Deep Learning for Infant Cry Recognition
Yun-Chia Liang1, Iven Wijaya1, Ming-Tao Yang2,3
1Department of Industrial Engineering and Management, Yuan Ze University, No. 135, Yuan-Tung Rd., Chung-Li Dist., Taoyuan City 32003, Taiwan.
Insights
This study uses deep learning (DL) to analyze infant cries, distinguishing between healthy and sick babies with 95% accuracy. It also identifies specific infant needs like hunger or pain, aiding parents in understanding their baby better.
Area of Science:
- Artificial Intelligence
- Infant Health
- Signal Processing
Background:
- Infant crying is a primary communication method, but its diverse causes are challenging for parents to interpret.
- Accurate identification of infant needs is crucial for well-being and parental support.
- Current methods for interpreting infant cries lack objective, data-driven approaches.
Purpose of the Study:
- To develop and evaluate deep learning (DL) models for recognizing infant needs and health status from audio recordings.
- To compare the performance of Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Artificial Neural Network (ANN) algorithms.
- To provide a technological solution for parents to better understand their infant's vocalizations.
Main Methods:
- Utilized 1607 audio recordings of infant cries, each 10 seconds long.
- Extracted audio features using mel-frequency cepstral coefficients (MFCC).
- Applied deep learning algorithms: CNN, LSTM, and ANN for classification tasks.
Main Results:
- CNN and LSTM achieved approximately 95% accuracy in differentiating healthy from sick infants.
- CNN demonstrated superior performance in recognizing specific infant needs (e.g., hunger, pain), reaching up to 60% accuracy.
- Both CNN and LSTM outperformed ANN in most performance metrics.
Conclusions:
- Deep learning models, particularly CNN, show significant potential in interpreting infant cries.
- These findings can inform the development of applications to assist parents in understanding infant communication.
- AI-powered analysis of infant vocalizations offers a promising avenue for improving infant care and parental support.
Abstract:
Recognizing why an infant cries is challenging as babies cannot communicate verbally with others to express their wishes or needs. This leads to difficulties for parents in identifying the needs and the health of their infants. This study used deep learning (DL) algorithms such as the convolutional neural network (CNN) and long short-term memory (LSTM) to recognize infants' necessities such as hunger/thirst, need for a diaper change, emotional needs (e.g., need for touch/holding), and pain caused by medical treatment (e.g., injection). The classical artificial neural network (ANN) was also used for comparison. The inputs of ANN, CNN, and LSTM were the features extracted from 1607 10 s audio recordings of infants using mel-frequency cepstral coefficients (MFCC). Results showed that CNN and LSTM both provided decent performance, around 95% in accuracy, precision, and recall, in differentiating healthy and sick infants. For recognizing infants' specific needs, CNN reached up to 60% accuracy, outperforming LSTM and ANN in almost all measures. These results could be applied as indicators for future applications to help parents understand their infant's condition and needs.
