Infant Distress Recognition & Soothing System

Infant Distress Recognition & Soothing System

Neonatal crying encodes information about hunger, discomfort, pain, or fatigue, but exhausted parents may struggle to interpret these signals promptly. This system employs a MEMS microphone to capture infant vocalizations, which are processed by a lightweight neural network classifier trained on labeled cry datasets. The model distinguishes among the primary cry etiologies with high accuracy, triggering context-appropriate responses: a gentle lullaby and rocking motion for sleep cues, or an alert to caregivers for pain indicators. The device mounts on the crib railing, maintaining safe distance while providing continuous acoustic monitoring. Privacy-preserving edge inference ensures no audio data leaves the local device.

Components

Raspberry Pi Zero
MEMS Microphone
Speaker Module
Vibration Motor
LED Indicator
Wi-Fi Module