🎭 Emotion Detector - Chrome Extension

A real-time emotion detection tool for video meetings. Captures frames from your video call and displays detected emotions as beautiful popups.

---

🚀 Quick Start (5 minutes)

Step 1: Install Dependencies

pip install fastapi uvicorn opencv-python pillow transformers torch

Step 2: Run the Backend API

Save your main.py file and run:

python main.py

You should see:

INFO:     Uvicorn running on http://127.0.0.1:8000

Keep this terminal open! The API needs to keep running.

Step 3: Install Chrome Extension

1. Go to chrome://extensions/
2. Turn on Developer mode (top-right toggle)
3. Click Load unpacked
4. Select your emotion-detector folder
5. A purple icon should appear in your toolbar

Step 4: Use the Extension

1. Open a video call (Google Meet, Teams, Zoom)
2. Click the purple icon
3. Enter API endpoint: http://127.0.0.1:8000/detect/file
4. Enter your name
5. Click Start Detection
6. Watch emotion popups appear! ✨

---

🔧 What's Running Where

Component	Location	Purpose
Backend API	http://127.0.0.1:8000	Analyzes images, returns emotions
Chrome Extension	In your browser toolbar	Captures video frames, sends to API, shows popups

---

📊 How It Works

1. You click "Start Detection" in the Chrome extension
2. Extension captures a frame from your video call every 1.5 seconds
3. Frame is sent to the API running on your computer
4. API analyzes the face and detects emotion (Happy, Sad, Angry, etc.)
5. Emotion popup appears on your screen in real-time
6. Data is logged for later export as JSON

---

💾 Export Your Data

1. Click Export Logs in the extension popup
2. A JSON file downloads with timestamps and emotions
3. Use it for analysis or reporting

Example log:

[
  {
    "timestamp": "2025-01-15T10:30:45.123Z",
    "participant_name": "You",
    "emotion": "HAPPY"
  },
  {
    "timestamp": "2025-01-15T10:30:46.500Z",
    "participant_name": "You",
    "emotion": "NEUTRAL"
  }
]

---

📋 System Requirements

• Python 3.8+
• Chrome/Chromium browser
• 4GB RAM (minimum)
• Internet connection (for first-time model download)

---

📝 Notes

• Emotions are detected locally on your computer (no cloud upload)
• Data is stored in browser storage, exportable as JSON
• Works on Google Meet, Microsoft Teams, Zoom, and any video site
• Keep the API terminal window open while using the extension

---

🚀 Next Steps

• Detect multiple people: Modify emotion_capture.py to detect all faces instead of just the largest one
• Real-time dashboard: Build a web dashboard to visualize emotions over time
• Custom emotions: Train the model with your own emotion labels
• Sound notifications: Add audio alerts when specific emotions are detected
• Cloud deployment: Deploy the API to Vercel, AWS, or Heroku for remote use
• Video recording: Save video clips with emotion timestamps