NSFAS Algorithm Lab

A showcase and testing environment for innovative algorithms and solutions designed for NSFAS. This platform demonstrates cutting-edge approaches to real-world challenges including face extraction, verification, and other machine learning solutions.

Architecture

Backend API (Hugging Face Space)

• Location: huggingface-space/
• Python-based face extraction API using FastAPI
• Deployed separately on Hugging Face Spaces
• Handles face detection, extraction, and comparison
• See huggingface-space/README.md for details

Frontend (Next.js)

• Location: src/
• React-based web interface
• Deployed on Vercel
• Connects to HuggingFace Space API
• Modern UI with Tailwind CSS + Framer Motion

Tech Stack

• Frontend: Next.js 14, React 18, Tailwind CSS, Framer Motion, Lucide Icons
• Backend: FastAPI, Face Recognition, OpenCV, PyMuPDF
• Deployment: Vercel (Frontend) + HuggingFace Spaces (Backend API)

Documentation

• FACE_EXTRACTION_API_MIGRATION.md - API migration guide
• FINAL_SETUP_INSTRUCTIONS.md - Setup instructions
• NEXT_STEPS.md - Development roadmap
• PREPROD_CHANGELOG.md - Change history
• PREPROD_IMPLEMENTATION.md - Implementation details
• PREPROD_README.md - Original preprod documentation
• PREPROD_STRUCTURE.md - Project structure
• PREPROD_THEME_GUIDE.md - Theme and styling guide

Quick Start

Install Dependencies

npm install

Development

npm run dev

Environment Variables

Create a .env.local file:

NEXT_PUBLIC_HF_FACE_API_URL=https://your-huggingface-space.hf.space

Deployment

• Frontend deploys automatically to Vercel via GitHub
• Backend API runs on HuggingFace Space

See FINAL_SETUP_INSTRUCTIONS.md for detailed setup instructions.

License

See LICENSE file in the main NSFAS repository.