Real-time emotion detection for mobile

A React Native library that detects emotions and tracks motion using on-device AI. No cloud, no servers, just your phone.

😊Happy - positive mood

🤩Excited - high energy

😴Tired - drowsiness detected

😐Neutral - baseline

For Developers

Use this library in your own React Native apps. All metrics are available through simple APIs.

emotion

Current emotional state: happy | excited | tired | neutral

confidence

Detection confidence level (0-100)

smileIntensity

Smile percentage (0.0 - 1.0)

headMovement

Head motion activity (0-100)

eyesClosed

Boolean indicating if eyes are closed

drowsinessLevel

Drowsiness percentage (0-100)

energyLevel

Overall activity level (0-100)

eyeClosureRate

Rate of eye closure (0.0 - 1.0)

EmotionDetector.initialize()

Initialize the emotion detection system

EmotionDetector.detectEmotion(faces)

Analyze faces and return emotion data

EmotionDetector.getStats()

Get aggregated statistics over time

EmotionDetector.reset()

Clear history and reset detectors

1.Face Detection - ML Kit's TensorFlow Lite model detects faces and facial landmarks on Arm processors
2.Smile Analysis - Measures smile probability with exponential smoothing to prevent sudden drops
3.Eye Tracking - Monitors eye openness probability and calculates closure patterns
4.Motion Detection - Tracks head position changes and calculates movement velocity
5.Emotion Classification - Custom algorithm combines all metrics to determine emotional state

React NativeExpoML Kit (TensorFlow Lite)TypeScriptVision Camera