science

Computer can tell if a baby is hungry, tired or in pain by listening to their cries


Scientists create an AI-powered computer system that can tell if a baby is hungry, tired or in pain by listening to their cries

  • The system was developed by the IEEE and the Chinese Automation Association
  • Scientists developed an AI system that can tell what babies’ cries actually mean
  • While each baby’s cry is unique, they share some features or sound signals
  • The researchers utilised this when training the AI through the differences between sound signals that actually carry the information

A computer programme, powered by artificial intelligence,  is capable of identifying and differentiating between babies’ cries to tell parents what their children want. 

The system, developed by researchers from the IEEE and the Chinese Association of Automation, can detect whether the cry is because of hunger, fatigue. illness or if it is in pain. 

While each baby’s cry is unique, they share some common features when they result from the same reasons. 

The team used an algorithm based on automatic speech recognition to detect and recognise the features of infant cries. 

Scroll down for video 

A new artificial intelligence method is capable of identifying and differentiating between babies' cries to tell what they want, or even if they're in pain. The system can detect whether the cry is because of hunger, fatigue or illness.

A new artificial intelligence method is capable of identifying and differentiating between babies’ cries to tell what they want, or even if they’re in pain. The system can detect whether the cry is because of hunger, fatigue or illness.

 These sound signals were identified using ‘compressed sensing’, a process that reconstructs a sound based on sparse data, especially useful for noisy environments.

THE MOST COMMON REASONS THAT BABIES CRY

The most common reasons for crying are: hunger 

– a dirty or wet nappy 

– tiredness 

– wanting a cuddle 

– wind 

– being too hot or too cold 

– boredom 

– overstimulation 

The researchers designed a cry language recognition algorithm which can distinguish the meanings of both normal and abnormal cry signals in a noisy environment. 

The algorithm is picks up the signals of the individual crier, and then stores that information and what the cry meant.

This means that it can be used in a broader sense in practical scenarios as a way to recognise and classify various cry features and better understand why babies are crying and how urgent the cries are. 

‘Like a special language, there is lots of health-related information in various cry sounds,’ said Lichuan Liu, corresponding author and Associate Professor of Electrical Engineering and the Director of Digital Signal Processing Laboratory.

‘The differences between sound signals actually carry the information. These differences are represented by different features of the cry signals. 

While each baby's cry is unique, they share some common features when they result from the same reasons. The team used an algorithm based on automatic speech recognition to detect and recognise the features of infant cries

While each baby’s cry is unique, they share some common features when they result from the same reasons. The team used an algorithm based on automatic speech recognition to detect and recognise the features of infant cries

‘To recognise and leverage the information, we have to extract the features and then obtain the information in it.’  

The team hope that the findings could be applicable to other medical care circumstances in which decision making relies heavily on experience. 

‘The ultimate goals are healthier babies and less pressure on parents and care givers,’ said Dr Liu. 

‘We are looking into collaborations with hospitals and medical research centers, to obtain more data and requirement scenario input, and hopefully we could have some products for clinical practice.’ 

The research was published in the journal of Automatica Sinica (JAS).



READ SOURCE

Leave a Reply

This website uses cookies. By continuing to use this site, you accept our use of cookies.