face10ai - Face Attractiveness Rating Platform

A modern SaaS application that analyzes facial attractiveness using AI, built with Next.js, TypeScript, and Tailwind CSS.

🚀 Features

• AI-Powered Face Analysis: Uses face-api.js for client-side face detection and analysis
• Consistent Results: Same photo always returns the same rating (SHA-256 hashing)
• Credit System: Free tier with 1 anonymous rating, 5 credits for registered users
• Referral Program: Earn 10 credits per successful referral
• Authentication: Email/password and Google OAuth via NextAuth.js
• Modern UI: Responsive design with Framer Motion animations
• SEO Optimized: Comprehensive metadata, sitemap, and Open Graph tags
• GDPR Compliant: Privacy-first approach, photos never stored
• French Language: All UI and content in French

📋 Prerequisites

Before you begin, ensure you have:

• Node.js 18+ installed
• PostgreSQL database
• Google OAuth credentials (optional, for Google login)

🛠️ Installation

1. Clone the repository

git clone <your-repo-url>
cd face10ai

2. Install dependencies

npm install

3. Set up environment variables

Create a .env.local file in the root directory:

# Database
DATABASE_URL="postgresql://user:password@localhost:5432/face10ai?schema=public"

# NextAuth
NEXTAUTH_URL=http://localhost:3000
NEXTAUTH_SECRET=your_nextauth_secret_here_generate_with_openssl_rand_base64_32
GOOGLE_CLIENT_ID=your_google_client_id
GOOGLE_CLIENT_SECRET=your_google_client_secret

# Application
NEXT_PUBLIC_APP_URL=http://localhost:3000
REFERRAL_CREDITS_AMOUNT=10
INITIAL_SIGNUP_CREDITS=5

# AttractiveNet API (optional - for AI beauty scoring)
NEXT_PUBLIC_PYTHON_API_URL=http://localhost:8000
NEXT_PUBLIC_USE_ATTRACTIVENET=true

4. Generate NextAuth secret

openssl rand -base64 32

5. Set up the database

npx prisma migrate dev --name init
npx prisma generate

6. Download face-api.js models

Create a public/models directory and download the required models from the face-api.js repository:

• tiny_face_detector_model-weights_manifest.json
• tiny_face_detector_model-shard1
• face_landmark_68_model-weights_manifest.json
• face_landmark_68_model-shard1
• face_recognition_model-weights_manifest.json
• face_recognition_model-shard1

Or use this command:

mkdir -p public/models
cd public/models
wget https://raw.githubusercontent.com/justadudewhohacks/face-api.js/master/weights/tiny_face_detector_model-weights_manifest.json
wget https://raw.githubusercontent.com/justadudewhohacks/face-api.js/master/weights/tiny_face_detector_model-shard1
wget https://raw.githubusercontent.com/justadudewhohacks/face-api.js/master/weights/face_landmark_68_model-weights_manifest.json
wget https://raw.githubusercontent.com/justadudewhohacks/face-api.js/master/weights/face_landmark_68_model-shard1
wget https://raw.githubusercontent.com/justadudewhohacks/face-api.js/master/weights/face_recognition_model-weights_manifest.json
wget https://raw.githubusercontent.com/justadudewhohacks/face-api.js/master/weights/face_recognition_model-shard1
cd ../..

7. Install Python dependencies (for AttractiveNet API)

If you want to use the AttractiveNet model for AI-powered beauty scoring:

⚠️ IMPORTANT: TensorFlow requires Python 3.8-3.11. Python 3.12+ is NOT supported.

If you have Python 3.12+ (like Python 3.14): 👉 See training/WINDOWS_SETUP.md for detailed setup instructions.

Quick start (if you have Python 3.11):

Windows:

cd training
py -3.11 -m venv venv
.\venv\Scripts\Activate.ps1
pip install -r requirements-windows.txt

macOS/Linux:

cd training
python3.11 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

See training/INSTALL_WINDOWS.md for detailed Windows installation instructions.

For macOS/Linux:

cd training
pip install -r requirements.txt

Note: If you encounter TensorFlow installation errors, you're likely using Python 3.12+. Install Python 3.11 and use it specifically.

8. Run the development server

In one terminal, start the Next.js app:

npm run dev

Open http://localhost:3000 with your browser to see the result.

Note: Face scoring now runs entirely in the browser using TensorFlow.js models (public/models/beauty_model_male and public/models/beauty_model_female). There is no geometric fallback and no separate AttractiveNet API process; if the AI models are missing or fail to load, scoring will be unavailable until this is fixed.

🏗️ Project Structure

face10ai/
├── app/                          # Next.js app directory
│   ├── api/                      # API routes
│   │   ├── auth/[...nextauth]/   # NextAuth configuration
│   │   ├── upload/               # Image upload endpoint
│   │   ├── analyze/              # Face analysis endpoint
│   │   ├── credits/              # Credit management
│   │   ├── referral/             # Referral tracking
│   │   └── signup/               # User registration
│   ├── auth/                     # Authentication pages
│   ├── dashboard/                # User dashboard
│   ├── result/[id]/              # Result display page
│   ├── about/                    # About page
│   ├── privacy/                  # Privacy policy
│   ├── terms/                    # Terms of service
│   ├── layout.tsx                # Root layout
│   ├── page.tsx                  # Homepage
│   ├── sitemap.ts                # Dynamic sitemap
│   └── robots.ts                 # Robots.txt
├── components/                   # React components
│   ├── Navbar.tsx
│   ├── Footer.tsx
│   ├── GenderSelector.tsx
│   ├── ImageUploader.tsx
│   ├── RatingDisplay.tsx
│   ├── CreditBadge.tsx
│   ├── ShareButtons.tsx
│   ├── ReferralSection.tsx
│   └── SessionProvider.tsx
├── lib/                          # Utility functions
│   ├── prisma.ts                 # Prisma client
│   ├── supabase.ts               # Supabase client
│   ├── auth.ts                   # Auth helpers
│   ├── credits.ts                # Credit management
│   ├── referrals.ts              # Referral system
│   ├── image-hash.ts             # Image hashing
│   ├── anonymous-tracking.ts     # Anonymous user tracking
│   └── face-rating.ts            # Face analysis algorithm
├── prisma/
│   └── schema.prisma             # Database schema
├── public/
│   └── models/                   # Face-api.js models
├── types/
│   └── next-auth.d.ts            # NextAuth type extensions
└── middleware.ts                 # Route protection

🎯 Key Features Implementation

Face Analysis Algorithm

The face rating algorithm uses face-api.js to:

1. Detect faces in uploaded images
2. Extract 68 facial landmarks
3. Calculate symmetry score (left vs right comparison)
4. Evaluate proportions based on golden ratio
5. Assess feature quality and spacing
6. Generate a deterministic score (1-10)

Credit System

• Anonymous users: 1 free rating
• New accounts: 5 credits
• Referrals: 10 credits per successful referral
• Credits are tracked per user with full transaction history

Caching & Consistency

• Images are hashed using SHA-256
• Results are cached in database by hash
• Same image always returns same score (from cache)

🔒 Security & Privacy

• Photos are never stored on servers
• Analysis happens client-side in browser
• Only analysis results are saved (not images)
• Passwords are hashed with bcrypt
• GDPR compliant
• CSRF protection via NextAuth

📱 Responsive Design

All pages are fully responsive with:

• Mobile-first approach
• Tailwind CSS breakpoints
• Touch-friendly interactions
• Optimized for all screen sizes

🎨 Animations

Smooth animations using Framer Motion:

• Page transitions
• Component entrance animations
• Hover effects
• Score reveal animations

🚀 Deployment

Vercel (Recommended)

1. Push your code to GitHub
2. Import project in Vercel
3. Add environment variables
4. Deploy

Database Migration

Before first deployment:

npx prisma migrate deploy

📝 Environment Variables for Production

Make sure to set all environment variables in your production environment:

• DATABASE_URL: Your production database connection string
• NEXTAUTH_URL: Your production domain (e.g., https://face10ai.com)
• NEXTAUTH_SECRET: Generate a new secret for production
• GOOGLE_CLIENT_ID & GOOGLE_CLIENT_SECRET: OAuth credentials
• NEXT_PUBLIC_APP_URL: Your production URL

🧪 Testing

To test the application:

1. Anonymous user flow: Upload image without account
2. Sign up flow: Create account with referral code
3. Credit system: Verify credits awarded/deducted
4. Referral system: Test referral link and credit rewards
5. Consistency: Upload same image twice, verify same score

📈 Analytics (Optional)

Consider adding:

• Google Analytics
• PostHog
• Plausible Analytics

🤝 Contributing

1. Fork the repository
2. Create your feature branch
3. Commit your changes
4. Push to the branch
5. Open a Pull Request

📄 License

This project is proprietary and confidential.

📞 Support

For support, email contact@face10ai.com

🎉 Acknowledgments

• face-api.js for face detection
• Next.js team for the amazing framework
• Vercel for hosting platform
• All contributors and users

---

Built with ❤️ using Next.js, TypeScript, and Tailwind CSS