Baby breakthrough: New AI tool can tell parents what their newborn’s cries mean

A team of US researchers claim to have created the Holy Grail for parents, nurses and pediatricians alike: a device which can decipher baby talk or, more specifically, their cries.

New parents can often be brought to their wits’ end trying to figure out whether their newborn is sick, tired, hungry, in need of a hug, or perhaps something more serious.

Thankfully, a team of US researchers has developed a new artificial intelligence method that can distinguish babies' cries from one another, as well as differentiate an individual baby's critical and non-critical cries from each other e.g. whether the child is in distress or just hungry.

Using a proprietary cry language algorithm, based on automatic speech recognition, the researchers can detect hidden patterns in the infant’s cry using a compressed sensing technique which reconstructs the initial raw cry signal based on sparse data input.

This is incredibly useful for analyzing the babies' cries, especially in noisy environments, which is where they are most likely to be recorded i.e. nursing wards, maternity hospitals, in a chaotic home environment as desperate parents try to remember where they left the baby powder etc.

The algorithm can classify and log various cry features which then provide insight into why babies are crying and how urgent their situation is, crucial information for parents and healthcare workers alike.

“Like a special language, there are lots of health-related information in various cry sounds. The differences between sound signals actually carry the information,” explained Lichuan Liu of the Digital Signal Processing Laboratory which conducted the research.

The ultimate goals are healthier babies and less pressure on parents and caregivers.

The research is published in the May issue of the Journal of Automatica Sinica (JAS).

Think your friends would be interested? Share this story!