Ai Baby Monitor
Overview
The AI Baby Sound & Mood Monitor is a 100% client-side, privacy-first web application that turns any laptop, tablet, or smartphone into an intelligent nursery sentinel using only its built-in microphone and a web browser.
Key Features
-
100% On-Device Privacy: Deep learning inference executes locally in the browser via TensorFlow.js and WebGL. Zero raw audio data is recorded or uploaded to the cloud.
-
Instant Distress Detection: Identifies infant cries, screams, and whimpers using Google's pre-trained YAMNet model. High distress triggers an immediate pulsing red alert and dispatches an instant Telegram push notification to the parent's phone.
-
Dual-Polarity Mood Tracking: Unlike traditional alarm-only monitors, the system actively detects baby laughter and giggles, transitioning to a calming green pulse to track daytime emotional wellness.
-
Real-Time DSP & Auditing: Features a custom client-side audio resampler (48 kHz to 16 kHz), live waveform visualization, and a timestamped timeline logging every state transition.
-
Zero Extra Hardware: Requires no expensive IoT monitors or subscriptions—accessible instantly across devices via a single web link.