Missing Person Face Recognition System

Face recognition system for identifying missing persons using InsightFace/ArcFace embeddings and FAISS vector search.

Stack: FastAPI, React 18, SQLite/PostgreSQL, Redis, FAISS, InsightFace

Quick Start

chmod +x setup.sh && ./setup.sh
uvicorn main:app --reload
cd frontend && npm run dev 

Access at http://localhost:5173 | Default: admin / admin123

API

Docs: http://localhost:8000/docs

Public: POST /api/v1/search
Auth: POST /api/v1/auth/login
Admin: CRUD on /api/v1/admin/missing-persons (JWT required)

Configuration

See .env.example. Key settings:

DATABASE_URL=sqlite:///./missing_persons.db  # or postgresql://...
REDIS_URL=redis://localhost:6379/0
SECRET_KEY=change-in-production
FACE_DETECTION_CONFIDENCE=0.9
DEFAULT_MATCH_THRESHOLD=0.6

For PostgreSQL: createdb missing_persons_db && alembic upgrade head

Development

pytest --cov=.
alembic revision --autogenerate -m "description"

Performance

End-to-end search: <3s (detection ~100-500ms, embedding ~50-100ms, FAISS <100ms for 50k vectors)

Troubleshooting

• Lower FACE_DETECTION_CONFIDENCE for difficult images
• InsightFace models auto-download on first use
• Check Redis is running: redis-server

License

MIT