Young Father Creates a Neural Network for Detecting Baby’s Hunger

Good news for parents. Caleb Olsen, a Haitian-American IT developer, has created a neural network system that notifies users when their babies are hungry even before they start crying. He is currently waiting for the patent to be approved.

According to Olsen, a baby will not cry as soon as it gets hungry, but will first indicate this with a series of signals, such as smacking of the lips and the refusal of a pacifier.

Olsen’s system uses a camera to monitor his young son’s actions, Google’s MediaPipe framework, and the corresponding video processing algorithms to detect these signs. Once it has detected enough signals, it sends a notification to Olsen’s smartphone.

The benefit that this brings is that parents can react more quickly and prevent the babies from crying. And of course, less crying means more sleep for parents. Win-win.