Meme Matcher - Real-time Facial Expression to Meme Matching

A real-time computer vision application that matches your facial expressions and hand gestures to famous internet memes using MediaPipe's face and hand detection.

What It Does

Point your webcam at yourself, make different facial expressions and hand gestures, and watch as the app finds the meme that best matches your expression in real-time! The matched meme appears side-by-side with your camera feed.

Features

• Real-time Face Detection: Uses MediaPipe Face Landmarker to track 478 facial landmarks
• Hand Gesture Detection: Tracks hand positions to distinguish similar expressions (e.g., Leo's cheers vs Disaster Girl's smirk)
• Advanced Expression Analysis:
  • Eye openness (surprise, wide eyes)
  • Eyebrow position (raised, furrowed)
  • Mouth shape (smiling, open, concerned)
  • Hand gestures (raised hands, fist pumps)
• Smart Matching Algorithm: Weighted similarity scoring with exponential decay for accurate matching
• 6 Iconic Memes: Carefully selected for diverse expressions and high detection quality

Installation

Prerequisites

• Python 3.11
• Webcam

Setup

1. Clone the repository:

git clone 
cd make_me_a_meme

2. Install dependencies:

pip install mediapipe opencv-python numpy

3. Run the application:

python3 main.py

The app now starts a local web panel and opens your browser automatically, so camera permission is handled by the browser instead of the terminal-hosted Python window.

VSCode Flow

If you want to run it in a VSCode-style workflow:

• Open the folder in VSCode
• Run the task Run Meme Matcher Web
• Or start the launch profile Meme Matcher Web

This keeps Python only as the local backend and leaves camera permission to the browser.

Web Panel Route

If you want to start only the local route without the launcher:

python3 web_panel.py

Then open:

http://127.0.0.1:8080

The browser will request camera permission and the panel will show the live camera feed plus the best meme match.

How to Use

1. Run python3 main.py

2. Your webcam will activate

3. Make different expressions and gestures:

   • Angry face → Angry Baby
   • Smirk (no hands) → Disaster Girl
   • Smirk + hand on chin → Gene Wilder
   • Smile + raised hand → Leonardo DiCaprio
   • Wide eyes/staring → Overly Attached Girlfriend
   • Happy + fist pump → Success Kid

4. Press 'q' to quit

How It Works

1. Face & Hand Detection

• Uses MediaPipe Face Landmarker (478 landmarks per face)
• Uses MediaPipe Hand Landmarker (21 landmarks per hand, up to 2 hands)
• Detects facial features and hand positions in real-time

2. Feature Extraction

For each frame, the app calculates:

• Eye features: Openness, symmetry
• Eyebrow features: Height, position relative to eyes
• Mouth features: Openness, width ratio, elevation
• Hand features: Number of hands, raised/lowered position
• Expression scores: Surprise, smile, concern, cheers

3. Similarity Matching

• Compares your features against pre-loaded meme features
• Uses weighted exponential decay scoring
• Higher weights for distinctive features (cheers score: 30 points, hand_raised: 25 points)
• Finds the best match and displays it alongside your video feed

Contributing

Feel free to:

• Add more memes (with hand gestures for better accuracy!)
• Improve the matching algorithm
• Enhance the UI
• Optimize performance

License

This project is for educational and entertainment purposes.

Credits

• MediaPipe: Google's ML framework for face and hand detection
• Meme Images: Fair use, iconic internet memes
• OpenCV: Open source computer vision library

Future Improvements

• Add more memes (target: 10-15)
• GUI for meme selection
• Save matched screenshots
• Expression history/statistics
• Multiple face support
• Custom meme upload via UI
• Performance optimizations
• Mobile app version

---