MAASTTTII

A real-time camera-based emoji display application that uses MediaPipe to detect your poses and facial expressions, then displays the corresponding emoji in a separate window.

Features

• Hand Detection: Raise a hand above your shoulder → displays hands up emoji 🙌
• Tongue Out Detection: Stick your tongue out → displays tongue emoji 😛
• Side Eye Detection: Turn your head sideways → displays side eye emoji 🙄
• Smile Detection: Show your teeth → displays smiling emoji 😊
• Default State: Straight face → displays neutral emoji 😐
• Real-time Processing: Live camera feed with instant emoji reactions

Detection Priority

1. Hands Up (highest priority) — overrides all facial expression detection
2. Tongue Out — detected when tongue tip extends past lower lip
3. Side Eye — detected when nose offset from face center exceeds threshold
4. Smiling — detected via brightness of the mouth region (teeth visible)
5. Straight Face — default when none of the above are triggered

Requirements

• Python 3.12
• A webcam
• Required Python packages (see requirements.txt)

Installation

pip install -r requirements.txt

Usage

python emoji_reactor3.py

What Happens

1. Run the script — your webcam starts immediately.
2. Two windows will open:
   • Camera Feed: Shows your live camera with the current detection state overlaid
   • Emoji Output: Displays the corresponding emoji based on your expression or pose
3. Controls:
   • Press q to quit the application
   • Raise a hand above your shoulder for the hands up emoji 🙌
   • Stick your tongue out for the tongue emoji 😛
   • Turn your head to trigger the side eye emoji 🙄
   • Show your teeth / smile for the smiling emoji 😊
   • Keep a straight face for the neutral emoji 😐

How It Works

The application uses two MediaPipe solutions running in parallel each frame:

1. Pose Detection — Monitors shoulder and wrist landmark positions to detect raised hands
2. Face Mesh Detection — Analyzes 468 facial landmarks to detect expressions:
   • Teeth/Smile: Measures average pixel brightness of the mouth bounding box region
   • Tongue Out: Measures the distance from tongue-tip landmark to lower lip midpoint
   • Side Eye: Measures horizontal offset of the nose tip from the midpoint between both outer eye corners

Customization

Adjusting Detection Sensitivity

Edit the threshold constants at the top of emoji_reactor3.py:

Constant	Default	Effect
SMILE_THRESHOLD	0.22	Mouth aspect ratio for smile detection
HEAD_TURN_THRESHOLD	0.04	Nose offset for side-eye trigger
TONGUE_THRESHOLD	0.04	Tongue-tip distance for tongue-out trigger
TEETH_BRIGHTNESS_THRESHOLD	120	Pixel brightness to detect visible teeth

Changing Emoji Images

Replace the image files in the project root with your own:

File	Emoji	State
smile.jpg	😊	Smiling with teeth
plain.png	😐	Straight / neutral face
air.jpg	🙌	Hands raised up
tongue.jpg	😛	Tongue out
sideye.jpeg	🙄	Head turned / side eye

Script Versions

File	Features
emoji_reactor.py	v1 — Smile + Hands Up only
emoji_reactor2.py	v2 — Added Side Eye + Tongue Out (ratio-based)
emoji_reactor3.py	v3 ✅ Latest — Improved smile via teeth brightness detection

Technical Details

• Uses OpenCV for camera capture and window display
• Uses MediaPipe Pose (33 landmarks) for hand/shoulder tracking
• Uses MediaPipe FaceMesh (468 landmarks) for facial expression detection
• Frames are converted to RGB before MediaPipe processing, then back to BGR for display

Dependencies

• opencv-python — Computer vision and camera capture
• mediapipe — Pose and Face Mesh landmark detection
• numpy — Numerical computing for landmark geometry

See requirements.txt for version constraints and requirements-lock.txt for pinned versions.

License

MIT License — see LICENSE for details.