Americans taught artificial intelligence to understand the "language" of babies, Journal of Automatica Sinica reported.
Lichuan Liu from the University of Northern Illinois at De Kalbe does not exclude that in the future the development will find its application in clinical practice. The algorithm created by scientists is capable of classifying babies' cries with high accuracy. In the course of working on artificial intelligence, it turned out that young children in crying use a certain set of sounds, which varies depending on what the child is trying to say (desire to sleep, eat, complain about a dirty diaper, and so on).
By analyzing the record of baby crying, scientists found that babies emit about a dozen signals, and their structure was common to all children. This means that artificial intelligence will not need to “adapt” to each individual child to decipher them.