Baby Whisperer

The Baby Whisperer is a revamped baby monitor that uses voice enabled technology to identify variable crying patterns for infants. TensorFlow was used to process convolutional neural networks of the Mel Frequency Cepstral Coefficient of the infant cries audio files, and categorize them with a predictive reason of crying. The baby's cries can be recorded from a device that will associate the crying to a reason with the help of the neural network. The caregiver will also receive an SMS message with the reason at the time of recording. Additionally, a web browser is used to display the analytics of this data including most common reason as well as how many times a day the child has cried.

Share this project: