Hamster Computer Vision

A real-time camera app that shows a matching hamster based on your face expressions and gestures.

Demo

How It Works

What you do	Hamster shown
Finger heart (thumb + index together)	heart.png
Close both eyes	eyesclosed.png
Stick tongue out	tongue.png
Shrug shoulders	shrugging.png
Point one finger	actually.png
Smile (happy face)	happy.png
Neutral face	neutral.png

Setup

Install dependencies

pip install deepface opencv-python mediapipe==0.10.9 protobuf==4.25.3

Run

python3 main.py

The first run will automatically download 3 MediaPipe model files into the models/ folder. This only happens once — after that it works fully offline.

Press Q to quit.

Project Structure

hamster_recognition/
├── main.py              # Main program
├── pictures/            # Hamster images
│   ├── heart.png
│   ├── happy.png
│   ├── tongue.png
│   ├── eyesclosed.png
│   ├── actually.png
│   ├── shrugging.png
│   └── neutral.png
└── models/              # MediaPipe models (auto-downloaded)
    ├── hand_landmarker.task
    ├── face_landmarker.task
    └── pose_landmarker_lite.task

Tech Stack

• OpenCV — camera feed
• MediaPipe — hand, face, and pose detection
• DeepFace — facial emotion recognition