“Alexa, why is my baby crying?”
Scientists say they have developed a new artificial intelligence (AI) tool that can reliably interpret the cries of babies. Using similar technology to Amazon’s Alexa assistant, it can tell parents whether their little one is hungry, needs changing or is in pain.
Crying babies demand attention, but for their anxious parents it’s often difficult to understand exactly what the problem might be.
Now a group of American researchers has developed a AI that can help. The team says the system can distinguish between normal cries caused by hunger or a desire for bodily contact, for example, and more unusual ones resulting from illnesses.
Parents are often able to interpret the cries of their babies (even if this sometimes requires a degree of noisy and traumatic trial and error).
This is because the cries of babies share common features which parents consciously and unconsciously pick up on.
Identifying the patterns woven into the cacophonous wails of infants has proved even more difficult for scientific researchers, but some of the latest artificial intelligence applications have come to the rescue.
The team’s findings have been published jointly by the Institute of Electrical and Electronics Engineers, and the Chinese Association of Automation.
Their system uses automatic speech recognition check, the same technology used in digital assistants from companies like Google and Amazon, to detect and interpret different features of the baby’s crying.
The team used a process known as compressed sensing – which reconstructs signals based on incomplete data – and applied it to recognise and classify features within the general din to understand what’s making the baby cry, and how urgently it needs to be addressed.
“Like a special language, there are lots of health-related information in various cry sounds,” said Lichuan Liu, an associate professor at Northern Illinois University, and the team’s corresponding author.
“The differences between sound signals actually carry the information, she said, adding “these differences are represented by different features of the cry signals. To recognise and use the information, we have to extract the features and then obtain the information in it.”
In addition to giving parents peace of mind, it’s hoped the team’s findings will help healthcare professionals. “The ultimate goals are healthier babies and less pressure on parents and caregivers,” says Liu.
“We are looking into collaborations with hospitals and medical research centres, and hopefully we could have some products for clinical practice. — dpa