Drowsy — Drowsiness & Emotion Detection App

Real-time drowsiness and emotion detection system using a phone camera, a FastAPI backend, and a React Native (Expo) mobile app with Firebase integration.

🔧 Prerequisites

Tool	Version	Install
Python	3.11.x	https://www.python.org/downloads/release/python-3119/ (check "Add to PATH")
Node.js	18+	https://nodejs.org/
Expo Go	Latest	Install on your phone from App Store / Google Play
DroidCam	Latest	App on phone + PC client from https://www.dev47apps.com/

Your phone and PC must be connected to the same WiFi network.

📋 Step-by-Step Setup

1. Clone the repo

git clone https://github.com/Blhasn-Sehli/Drowsy.git
cd Drowsy

2. Find your PC's local IP address

You'll need this IP for both the backend camera and the frontend API connection.

Windows:

ipconfig

Look for Wireless LAN adapter WiFi → IPv4 Address (e.g. 192.168.1.XX)

macOS / Linux:

ifconfig | grep "inet "

Write down your IP, you'll use it in steps 4 and 5 below.

3. Setup Firebase credentials

Backend (Back/)

You need a firebaseKey.json file in the Back/ folder. Ask the project owner to send it to you, or create your own:

Go to Firebase Console → Project Settings → Service Accounts
Click "Generate new private key"
Save the file as Back/firebaseKey.json

A template is provided at Back/firebaseKey.example.json for reference.

Frontend (front/)

Create a file front/.env with your Firebase config:

EXPO_PUBLIC_FIREBASE_API_KEY=your_firebase_api_key
EXPO_PUBLIC_FIREBASE_AUTH_DOMAIN=your_project.firebaseapp.com
EXPO_PUBLIC_FIREBASE_DATABASE_URL=https://your_project-default-rtdb.firebaseio.com
EXPO_PUBLIC_FIREBASE_PROJECT_ID=your_project_id
EXPO_PUBLIC_FIREBASE_STORAGE_BUCKET=your_project.firebasestorage.app
EXPO_PUBLIC_FIREBASE_MESSAGING_SENDER_ID=your_sender_id
EXPO_PUBLIC_FIREBASE_APP_ID=your_app_id

Ask the project owner for the real values, or get them from Firebase Console → Project Settings → General → Your apps.

4. Configure the camera source (Backend)

Edit Back/config.py and set your phone's DroidCam IP:

CAMERA_INDEX: str = "http://YOUR_PHONE_IP:4747/video"

Example: if DroidCam shows 192.168.1.50, set:

CAMERA_INDEX: str = "http://192.168.1.50:4747/video"

Or use "0" to use your PC's built-in webcam.

5. ⚠️ Configure the API URL (Frontend) — IMPORTANT

Edit front/services/api.ts and change the IP to your PC's IP (from step 2):

const API_BASE_URL = 'http://YOUR_PC_IP:8000';

Example: if your PC's IP is 192.168.1.25:

const API_BASE_URL = 'http://192.168.1.25:8000';

This is the most common issue! The app won't connect if this IP doesn't match your PC.

6. Start the Backend

cd Back

Windows: Double-click setup.bat (first time only), then run.bat

Or manually:

python -m venv venv
source venv/Scripts/activate    # Windows
# source venv/bin/activate      # macOS/Linux
pip install -r requirements.txt
python main.py

You should see:

✅ Server: http://0.0.0.0:8000
✅ Continuous detection started

Test it: open http://localhost:8000/health in your browser.

7. Start the Frontend

cd front
npm install
npm start

Then scan the QR code with Expo Go on your phone.

🔍 Quick Checklist

Python 3.11 installed
Node.js 18+ installed
Phone and PC on the same WiFi
Back/firebaseKey.json exists with real credentials
front/.env exists with Firebase config
Back/config.py → camera IP matches your phone's DroidCam IP
front/services/api.ts → API URL matches your PC's WiFi IP
Backend running (python main.py)
Frontend running (npm start)

❌ Troubleshooting

Problem	Fix
Frontend can't connect to backend	Make sure the IP in `front/services/api.ts` matches your PC's WiFi IP (`ipconfig`)
Camera not working	Check `Back/config.py` — DroidCam IP must be correct, and phone + PC on same WiFi
Firebase errors	Make sure `Back/firebaseKey.json` and `front/.env` have real credentials
`python` not found	Reinstall Python 3.11 with "Add to PATH" checked
Port 8000 in use	Kill the process using it, or change the port in `Back/main.py`
Expo app can't connect	Make sure your phone is on the same WiFi as your PC

📁 Project Structure

Drowsy/
├── Back/                  # Python FastAPI backend
│   ├── main.py            # Server entry point
│   ├── detector.py        # Drowsiness detection (EAR/MAR)
│   ├── emotion_detector.py# Emotion detection (DeepFace)
│   ├── firebase_service.py# Firebase integration
│   ├── config.py          # Camera URL & thresholds ← EDIT THIS
│   ├── firebaseKey.json   # Firebase credentials (not in git)
│   └── requirements.txt   # Python dependencies
│
├── front/                 # React Native (Expo) app
│   ├── app/               # Screens (tabs)
│   ├── components/        # UI components
│   ├── services/
│   │   ├── api.ts         # Backend API connection ← EDIT THIS
│   │   └── firebase.ts    # Firebase realtime listener
│   ├── hooks/             # Custom React hooks
│   ├── .env               # Firebase config (not in git)
│   └── package.json
│
└── README.md              # ← You are here

Blhasn-Sehli/Drowsy